Skip to main content

Full text of "The Radio Handbook (William Orr, Editor) 15th Edition 1959"

See other formats 


This  book  is  revised  and  brought  up 
to  date  (at  irregular  intervals)  as 
necessitated  by  technical  progress. 

THE 
R Oil) 
HHDBOOK 


Fifteenth  Edition 


The  Standard  of  the  Field  — 

for  aduanced  amateurs 
practical  radiomen 
practical  engineers 
practical  technicians 


WILLIAM  I.  ORR,  W6SAI 

Editor,  15th  Edition 


$7.50  per  copy  ot 
your  dealer  in  U.S.A. 

(Add  10%  on  direct 
orders  to  publisher) 


Published  and  Distributed  to  the  Radio  Trade  by 

SoLctcy^  CLTbci 

L I M ' 1 e.  o O 


SUMMERLAND.  CALIFORNIA.  U.S.A. 


(Distributed  to  the  Book  and  News  Trades  and  Libraries  by  the  Baker  & Taylor  Co.,  Hillside,  N.  J.) 


THE  RADIO  HANDBOOK 


FIFTEENTH  EDITION 


Copyright,  1959,  by 

Editors  and  Engineers,  Ltd. 
Summerland,  Californio,  U.S.A. 

Copyright  under  Pan-American  Convention 
All  Translation  Rights  Reserved 


Printed  in  U.S.A. 


The  ''Radio  Handbook"  in  Sponish  or  Italian  is  available  from  us  ot  $8.25 
postpaid.  French  and  Dutch  editions  in  preporotion. 

Outside  North  America,  if  more  convenient,  write:  (Spanish)  Marcombo,  S.A.,  Av. 
Jose  Antonio,  584,  Barcelona,  Spain;  (Italion)  Edizione  C.E.L.I.,  Via  Gandino  1, 
Bologna,  Italy;  (French  or  Dutch)  P.  H.  Brans,  Ltd.,  28  Prins  Leopold  St.,  Borgerhout, 

Antwerp,  Belgium. 


Other  Outstanding  Books  from  the  Same  Publisher 
(See  Announcements  at  Back  of  Book) 

The  Radiotelephone  License  Manual 

The  Surplus  Radio  Conversion  Manuals 

The  World’s  Radio  Tubes  ( Radio  Tube  Vade  Mecum  ) 

The  World’s  Equivalent  Tubes  (Equivalent  Tube  Vade  Mecum) 

The  World’s  Television  Tubes  ( Television  Tube  Vade  Mecum  ) 


THE  RADIO  HANDBOOK 

15th  Edition 

Table  of  Contents 

ChopJer  One.  INTRODUCTION  TO  RADIO 11 

1-1  Amateur  Radio  

1-2  Station  and  Operotor  Licenses  12 

1-3  The  Amateur  Bands  12 

1- 4  Starting  Your  Study  14 

Chapter  Two.  DIRECT  CURRENT  CIRCUITS 21 

2- 1  The  Atom  21 

2-2  Fundamental  Electrical  Units  and  Relationships 22 

2-3  Electrostatics  — Copocitors  30 

2-4  Magnetism  ond  Electromagnetism  35 

2- 5  RC  and  RL  Transients  38 

Chapter  Three.  ALTERNATING  CURRENT  CIRCUITS 41 

3- 1  Alternating  Current  41 

3-2  Resonant  Circuits  53 

3-3  Nonsinusoldol  Waves  and  Transients  58 

3-4  Transformers  61 

3- 5  Electric  Filters  63 

Chapter  Four.  VACUUM  TUBE  PRINCIPLES 67 

4- 1  Thermionic  Emission  67 

4-2  The  Diode  71 

4-3  The  Triode  72 

4-4  Tetrode  or  Screen  Grid  Tubes  77 

4-5  Mixer  and  Converter  Tubes  79 

4-6  Electron  Tubes  at  Very  High  Frequencies  80 

4-7  Special  Microwave  Electron  Tubes  81 

4-8  The  Cathode-Roy  Tube  84 

4-9  Gas  Tubes  87 

4- 10  Miscellaneous  Tube  Types  88 

Chapter  Five.  TRANSISTORS  AND  SEMI-CONDUCTORS 90 

5- 1  Atomic  Structure  of  Germanium  and  Silicon  90 

5-2  Mechanism  of  Conduction  90 

5-3  The  Transistor  92 

5-4  Transistor  Characteristics  94 

5-5  Transistor  Circuitry  96 

5-6  Transistor  Circuits  103 


3 


Chapter  Six.  VACUUM  TUBE  AMPLIFIERS 106 

6-1  Vacuum  Tube  Parameters  106 

6-2  Classes  and  Types  of  Vacuum-Tube  Amplifiers  107 

6-3  Biasing  Methods  108 

6-4  Distortion  in  Amplifiers  109 

6-5  Resistance-Capacitance  Coupled  Audio-Frequency  Amplifiers....  109 

6-6  Video-Frequency  Amplifiers  113 

6-7  Other  Interstage  Coupling  Methods  113 

6-8  Phase  Inverters  115 

6-9  D-C  Amplifiers  117 

6-10  Single-ended  Triode  Amplifiers  118 

6-11  Single-ended  Pentode  Amplifiers  120 

6-12  Push-Pull  Audio  Amplifiers  121 

6-13  Class  B Audio  Frequency  Power  Amplifiers  123 

6-14  Cathode-Follower  Power  Amplifiers  127 

6-15  Feedback  Amplifiers  129 

6- 16  Vacuum-Tube  Voltmeters  130 

Chapter  Seven.  HIGH  FIDELITY  TECHNIQUES 134 

7- 1  The  Nature  of  Sound  134 

7-2  The  Phonograph  136 

7-3  The  High  Fidelity  Amplifier  138 

7-4  Amplifier  Construction  142 

7-5  The  “Baby  Hi  Fi’*  143 

7- 6  A High  Quality  25  Watt  Amplifier  146 

Chapter  Eight.  RADIO  FREQUENCY  VACUUM  TUBE  AMPLIFIERS 149 

Tuned  RF  Vacuum  Tube  Amplifiers  149 

8- 1  Grid  Circuit  Considerations  149 

8-2  Plate-Circuit  Considerations  151 

Radio-Frequency  Power  Amplifiers  152 

8-3  Class  C R-F  Power  Amplifiers  1 52 

8-4  Class  B Radio  Frequency  Power  Amplifiers  1 57 

8-5  Special  R-F  Power  Amplifier  Circuits  160 

8-6  A Grounded-Grid  304TL  Amplifier  163 

8- 7  Class  AB1  Radio  Frequency  Power  Amplifiers  165 

Chapter  Nine.  THE  OSCILLOSCOPE 170 

9- 1  A Typical  Cathode-Ray  Oscilloscope  170 

9-2  Display  of  Waveforms  175 

9-3  Lissajous  Figures  176 

9-4  Monitoring  Transmitter  Performance  with  the  Oscilloscope  179 

9-5  Receiver  l-F  Alignment  with  an  Oscilloscope  180 

9-6  Single  Sideband  Applications  182 

Chapter  Ten.  SPECIAL  VACUUM  TUBE  CIRCUITS 185 

10-1  Limiting  Circuits  185 

10-1  Clamping  Circuits  187 

10-3  Multivibrators  188 

10-4  The  Blocking  Oscillator  190 

10-5  Counting  Circuits  190 

10-6  Resistance  - Capacity  Oscillators  191 

10-7  Feedback  192 


4 


Chapter  Eleven.  ELECTRONIC  COMPUTERS 194 

11-1  Digital  Computers  195 

11-2  Binary  Notation  195 

11-3  Analog  Computers  197 

11-4  The  Operational  Amplifier  199 

11-5  Solving  Analog  Problems  200 

11-6  Non-linear  Functions  202 

11- 7  Digital  Circuitry  204 

Chapter  Tv/elve,  RADIO  RECEIVER  FUNDAMENTALS 207 

12- 1  Detection  or  Demodulation  207 

12-2  Superregenerative  Receivers  209 

12-3  Superheterodyne  Receivers  210 

12-4  Mixer  Noise  and  Images  212 

12-5  R-F  Stages  213 

12-6  Signal-Frequency  Tuned  Circuits  216 

12-7  l-F  Tuned  Circuits  218 

12-8  Detector,  Audio,  and  Control  Circuits  225 

12-9  Noise  Suppression  227 

12-10  Special  Considerations  in  U-H-F  Receiver  Design  231 

12-11  Receiver  Adjustment  235 

12- 12  Receiving  Accessories  236 

Chapter  Thirteen.  GENERATION  OF  RADIO  FREQUENCY  ENERGY 239 

13- 1  Self-Controlled  Oscillators  239 

13-2  Quartz  Crystal  Oscillators  244 

13-3  Crystal  Oscillator  Circuits  247 

13-4  Radio  Frequency  Amplifiers  251 

1 3-5  Neutralization  of  R.F.  Amplifiers  252 

13-6  Neutralizing  Procedure  255 

13-7  Grounded  Grid  Amplifiers  258 

13-8  Frequency  Multipliers  258 

13-9  Tank  Circuit  Capacitances  261 

13-10  L and  Pi  Matching  Networks  265 

13-11  Grid  Bios  267 

13-12  Protective  Circuits  for  Tetrode  Transmitting  Tubes  269 

13-13  Interstage  Coupling  270 

13-14  Radio-Frequency  Chokes  272 

13- 15  Porallel  and  Push-Pull  Tube  Circuits  273 

Chapter  Fourteen.  R-F  FEEDBACK 274 

14- 1  R-F  Feedback  Circuits  274 

14-2  Feedback  and  Neutralization  of  a Two-Stage  R-F  Amplifier....  277 

14- 3  Neutralization  Procedure  in  Feedback-Type  Amplifiers  279 

Chapter  Fifteen.  AMPLITUDE  MODULATION 282 

15- 1  Sidebands  282 

15-2  Mechanics  of  Modulation  283 

15-3  Systems  of  Amplitude  Modulation  285 

15-4  Input  Modulation  Systems  292 

1 5-5  Cathode  Modulation  297 

15-6  The  Doherty  and  the  Terman-Woodyard  Modulated  Amplifiers..  298 

15-7  Speech  Clipping  300 

15-8  The  Bias-Shift  Heising  Modulator  307 


5 


Chapter  Sixteen.  FREQUENCY  MODULATION  AND  REDIOTELETYPE 

TRANSMISSION  312 

16-1  Frequency  Modulation  312 

16-2  Direct  FM  Circuits  315 

16-3  Phase  Modulation  319 

16-4  Reception  of  FM  Signals  321 

16- 5  Radio  Teletype  326 

Chapter  Seventeen.  SIDEBAND  TRANSMISSION 327 

17- 1  Commercial  Applications  of  SSB  327 

17-2  Derivation  of  Single-Sidebond  Signals  328 

17-3  Carrier  Elimination  Circuits  332 

1 7-4  Generation  of  Single-Sideband  Signals  334 

17-5  Single  Sideband  Frequency  Conversion  Systems  340 

17-6  Distortion  Products  Due  to  Nonlinearity  of  R-F  Amplifiers  344 

17-7  Sideband  Exciters  346 

17-8  Reception  of  Single  Sideband  Signals  351 

17- 9  Double  Sideband  Transmission  353 

Chapter  Eighteen.  TRANSMITTER  DESIGN 356 

18- 1  Resistors  356 

18-2  Capacitors  358 

1 8-3  Wire  and  Inductors  360 

1 8-4  Grounds  362 

1 8-5  Holes,  Leads  and  Shafts 362 

1 8-6  Parasitic  Resonances  364 

18-7  Parasitic  Oscillation  in  R-F  Amplifiers  365 

1 8-8  Elimination  of  V-H  F Parasitic  Oscillations  366 

18- 9  Checking  for  Parasitic  Oscillations  368 

Chapter  Nineteen.  TELEVISION  AND  BROADCAST  INTERFERENCE 371 

19- 1  Types  of  Television  Interference  371 

19-2  Harmonic  Radiation  373 

19-3  Low-Pass  Filters  376 

19-4  Broadcast  Interference  379 

19- 5  HI-FI  Interference  386 

Chapter  Twenty.  TRANSMITTER  KEYING  AND  CONTROL 387 

20- 1  Power  Systems  387 

20-2  Transmitter  Control  Methods  391 

20-3  Safety  Precautions  393 

20-4  Transmitter  Keying  395 

20-5  Cathode  Keying  397 

20-6  Grid  Circuit  Keying  398 

20-7  Screen  Grid  Keying  399 

20- 8  Differentiai  Keying  Circuits  400 

Chapter  Twenty-One.  RADIATION,  PROPAGATION  AND  TRANSMISSION 

LINES  403 

21- 1  Radiation  from  an  Antenna  403 

21-2  General  Characteristics  of  Antennas  404 

21-3  Radiation  Resistance  and  Feed-Point  Impedance  407 

21-4  Antenna  Directivity  410 

21-5  Bandwidth  413 


6 


21-6  Propagation  of  Radio  Waves  413 

21-7  Ground- Wave  Communication  414 

21-8  Ionospheric  Propagation  416 

21-9  Transmission  Lines  420 

21-10  Non-Resonant  Transmission  Lines  421 

21-11  Tuned  or  Resonant  Lines  424 

21- 12  Line  Discontinuities  425 

Chapter  Twenty-Two.  ANTENNAS  AND  ANTENNA  MATCHING 426 

22- 1  End-Fed  Half-Wave  Horizontal  Antennas  426 

22-2  Center-Fed  Half-Wave  Horizontal  Antennas  427 

22-3  The  Half-Wave  Vertical  Antenna  430 

22-4  The  Ground  Plane  Antenno  431 

22-5  The  Marconi  Antenna  432 

22-6  Space-Conserving  Antennas  434 

22-7  Multi-Band  Antennas  436 

22-8  Matching  Non-Resonont  Lines  to  the  Antenna  442 

22-9  Antenna  Construction  448 

22-10  Coupling  to  the  Antenna  System  451 

22-11  Antenna  Couplers  454 

22- 12  A Single-Wire  Antenna  Toner 456 

Chapter  Twenty-Three.  HIGH  FREQUENCY  ANTENNA  ARRAYS 459 

23- 1  Directive  Antennas  459 

23-2  Long  Wire  Radiators  461 

23-3  The  V Antenna  462 

23-4  The  Rhombic  Antenna  464 

23-5  Stacked-Dipole  Arroys  465 

23-6  Broadside  Arrays  468 

23-7  End-Fire  Directivity  473 

23- 8  Combination  End-Fire  and  Broadside  Arrays  475 

Chapter  Twenty-Four.  V-H-F  AND  U-H-F  ANTENNAS  477 

24- 1  Antenna  Requirements  477 

24-2  Simple  Horizontally-Polarized  Antennas  479 

24-3  Simple  Vertical-Polarized  Antennas  480 

24-4  The  Discone  Antenno  481 

24-5  Helical  Beam  Antennas  483 

24-6  The  Corner-Reflector  and  Horn-Type  Antennas  485 

24-7  VHF  Horizontal  Rhombic  Antenna  486 

24- 8  Multi-Element  V-H-F  Beam  Antennas  488 

Chapter  Twenty-Five.  ROTARY  BEAMS 494 

25- 1  Unidirectional  Parasitic  End-Fire  Arrays  (Yogi  Type)  494 

25-2  The  Two  Element  Beam  494 

25-3  The  Three-Element  Array  496 

25-4  Feed  Systems  for  Parasitic  (Yogi)  Arrays  498 

25-5  Unidirectional  Driven  Arrays  504 

25-6  Bi-Directional  Rotatable  Arrays  505 

25-7  Construction  of  Rotatable  Arrays  506 

25-8  Tuning  the  Array  509 

25-9  Antenna  Rotation  Systems  513 

25-10  Indication  of  Direction  514 

25-11  “Three-Bands**  Beams  514 


7 


Chapter  Twenty-Six.  MOBILE  EQUIPMENT  DESIGN  AND  INSTALLATION....51 5 

26-1  Mobile  Reception  515 

26-2  Mobile  Transmitters  521 

26-3  Antennas  for  Mobile  Work  522 

26-4  Construction  and  Installation  of  Mobile  Equipment  524 

26- 5  Vehicular  Noise  Suppression  527 

Chapter  Twenty-Seven.  RECEIVERS  AND  TRANSCEIVERS 530 

27- 1  Circuitry  and  Components  533 

27-2  A Simple  Transistorized  Portable  B-C  Receiver  533 

27-3  A 455  Kc.  Mechanical  Filter  Adapter  535 

27-4  A High  Performance  Amateur  Band  Receiver  540 

27-5  A “Handie-Talkie  * for  144  Me 547 

27-6  Six  Meter  Transceiver  for  Home  or  Car  552 

27- 7  A “Hot”  Transceiver  for  28  Megacycles  559 

Chapter  Twenty-Eight.  LOW  POWER  TRANSMITTERS  AND  EXCITERS 567 

28- 1  SSB  Exciter  for  Fixed  or  Mobile  Use  567 

28-2  A Mobile  Transistorized  SSB  Exciter  574 

28-3  A VHF  Transceiver  of  Advanced  Design  578 

28-4  A Miniaturized  SSB  Transmitter  for  14  Me 589 

28-5  A Duplex  Transmitter-Receiver  for  220  Me 598 

28- 6  A High  Stability  V.F.O.  For  the  DX  Operator  604 

Chapter  Twenty-Nine.  HIGH  FREQUENCY  POWER  AMPLIFIERS 610 

29- 1  Power  Amplifier  Design  610 

29-2  Push-Pull  Triode  Amplifiers  612 

29-3  Push-Pull  Tetrode  Amplifiers  614 

29-4  Tetrode  Pi-Network  Amplifiers  617 

29-5  A Compact  Linear  Amplifier  for  Mobile  SSB  620 

29-6  A Multi-band  Mobile  Linear  Amplifier  624 

29-7  An  Inexpensive  Cathode  Driven  Kilowatt  Amplifier  626 

29-8  A Low  Distortion  Sideband  Linear  Amplifier  629 

29-9  Kilowatt  Amplifier  for  Linear  or  Class  C Operation  635 

29-10  A 2 Kilowatt  P.E.P,  All-band  Amplifier  640 

29- 1 1 A High  Power  Push-pull  Tetrode  Amplifier  644 

Chapter  Thirty.  SPEECH  AND  AMPLITUDE  MODULATION  EQUIPMENT 647 

30- 1  Modulation  647 

30-2  Design  of  Speech  Amplifiers  and  Modulators  650 

30-3  General  Purpose  Triode  Closs  B Modulator  651 

30-4  A 10-Watt  Amplifier-Driver  655 

30-5  500-Watt  304TL  Modulator  656 

30-6  A 15-Watt  Clipper-Amplifier  657 

30-7  A 200-Watt  Sll-A  De-Luxe  Modulator  658 

30- 8  Zero  Bias  Tetrode  Modulators  662 

Chapter  Thirty-One.  TRANSMITTER  CONSTRUCTION 663 

31- 1  A 300  Watt  Phone/C-W  Transmitter  for  50/144  Me 663 

31-2  A De-Luxe  Transmitter  for  the  3.5-29.7  Me.  Range  673 


8 


Chapter  Thirty-Two.  POWER  SUPPUES  684 

32-1  Power  Supply  Requirements  684 

32-2  Rectification  Circuits  689 

32-3  Standard  Power  Supply  Circuits  690 

32-4  Selenium  and  Silicon  Rectifiers  695 

32-5  100  Watt  Mobile  Power  Supply  697 

32-6  Transistorized  Power  Supplies  703 

32-7  Two  Transistorized  Mobile  Supplies  706 

32-8  Power  Supply  Components  707 

32-9  Special  Power  Supplies  709 

32-10  Power  Supply  Design  712 

32-11  300  Volt,  50  Ma.  Power  Supply  715 

32-12  500  Volt,  200  Milliompere  Power  Supply  716 

32-13  1500  Volt,  425  Milliampere  Power  Supply  717 

32-14  A Dual  Voltage  Transmitter  Supply  718 

32- 15  A Kilowatt  Power  Supply  718 

Chapter  Thirty-Three.  WORKSHOP  PRACTICE 720 

33- 1  Tools  720 

33-2  The  Material 723-A 

33-3  TVI-Proof  Enclosures 724-A 

33-4  Enclosure  Openings 725-A 

33-5  Summation  of  the  Problem 725-A 

33-6  Construction  Practice 726-A 

33- 7  Shop  Layout 729-A 

Chapter  Thirty-Four.  ELECTRONIC  TEST  EQUIPMENT 721-B 

34- 1  Voltage,  Current  and  Power 721-B 

34-2  Measurement  of  Circuit  Constants 727-B 

34-3  Measurements  with  a Bridge 728-B 

34-4  Frequency  Measurements 729-B 

34-5  Antenna  and  Transmission  Line  Measurements 730-B 

34-6  A Simple  Coaxial  Reflectometer  732 

34-7  Measurements  on  Balonced  Transmission  Lines  734 

34-8  A **Balanced"  SWR  Bridge  736 

34-9  The  Antennascope  733 

34-10  A Silicon  Crystal  Noise  Generator  740 

Chapter  Thirty-Five.  RADIO  MATHEMATICS  AND  CALCULATIONS 742 


9 


FOREWORD  TO  THE  FIFTEENTH  EDITION 


Over  two  decades  ago  the  historic  first  edition  of  the  RADIO  HANDBOOK  was 
published  as  a unique,  independent,  communications  manual  written  especially  for  the 
advmced  radio  amateur  and  electronic  engineer.  Since  that  early  issue,  great  pains  have 
been  taken  to  keep  each  succeeding  edition  of  the  RADIO  HANDBOOK  abreast  of 
the  rapidly  expanding  field  of  electronics. 

So  quickly  has  the  electron  invaded  our  everyday  affairs  that  it  is  now  no  longer 
possible  to  segregate  one  particular  branch  of  electronics  and  define  it  as  radio  com- 
munications; rather,  the  transfer  of  intelligence  by  electrical  means  encompasses  more 
than  the  vacuum  tube,  the  antenna,  and  the  tuning  capacitor. 


Included  in  this  new,  advanced  Fifteenth  Edition  of  the  RADIO  HANDBOOK  are 
fresh  chapters  covering  electronic  computers,  r.f.  feedback  amplifiers,  and  high  fidelity 
techniques,  plus  greatly  expanded  chapters  dealing  with  semi-conductors  and  special 
vacuum  tube  circuits.  The  other  chapters  of  this  Handbook  have  been  thoroughly 
revised  and  brought  up  to  date,  touching  briefly  on  those  aspects  in  the  industrial 
and  military  electronic  fields  that  are  of  immediate  interest  to  the  electronic  engineer 
and  the  radio  amateur.  The  construction  chapters  have  been  completely  re-edited.  All 
new  equipments  described  therein  are  of  modern  design,  free  of  TVI  problems  and 
various  unwanted  parasitic  oscillations.  An  attempt  has  been  made  not  to  duplicate 
items  that  have  been  featured  in  contemporary  magazines.  The  transceiver  makes  its 
major  bow  in  this  edition  of  the  RADIO  HANDBOOK,  and  it  is  felt  that  this 
complete,  inexpensive,  compact  "radio  station"  design  will  become  more  popular 
during  the  coming  years. 

The  writing  and  preparation  of  this  Handbook  would  have  been  impossible  without 
the  lavish  help  that  was  tended  the  editor  by  fellow  amateurs  and  sympathetic  elec- 
tronic organizations.  Their  friendly  assistance  and  helpful  suggestions  were  freely 
given  in  the  true  amateur  spirit  to  help  make  the  1 5th  edition  of  the  RADIO  HAND- 
BOOK an  outstanding  success. 


The  editor  and  publisher  wish  to  thank  these  individuals  and  companies  whose 
unselfish  support  made  the  compilation  and  publication  of  this  book  an  interesting 
and  inspired  task.  — ^William  I.  Orr,  W6SAI,  3A2AF,  Editor 


E.  P.  Alvernaz,  W6DMN, 
Jennings  Radio  Co. 

Kenneth  Bay,  W2GSJ, 

General  Electric  Co. 

Orrin  H.  Brown,  W6HB, 
Eitel-McCullough,  Inc. 

Wm.  E.  Bruring,  W9Z,SO, 

E.  F.  Johnson,  Inc. 

Thomas  Consalvi,  W3EOZ, 
Barker  & Williamson,  Inc. 

Cal  Hadlock,  WICTW, 
National  Co.,  Inc. 

Jo  E.  Jennings,  W6EI, 
Jennings  Radio  Co. 

A1  Kahn,  W8DUS, 
Electrovoice,  Inc. 

Ken  Klippel,  WOSQO, 

Collins  Radio  Co. 

Roger  Mace,  W8MWZ, 

Heath  Co. 

E.  R.  Mullings,  W8VPN, 
Heath  Co. 

Edw.  A.  Neal,  W2JZK, 
General  Electric  Co. 

Edw.  Schmeichel,  W9YFV, 
Chicago-Standard 
Transformer  Co. 


Wesley  Schum,  W9DYV, 

Central  Electronics,  Inc. 

Aaron  Self,  W8FYR, 

Continental  Electronics 
& Sound  Co. 

Harold  Vance,  K2FF, 

Radio  Corporation  of 
America 

J.  A.  Haimes,  Semi-conductor 
Division,  Radio  Corporation 
of  America 

Special  thanks  are  due  Collins 
Radio  Co.  for  permission  to 
reprint  portions  of  their 
Sideband  Report  CTR-113 
by  Warren  Bruene,  WOTTK 

Bud  Radio  Co.,  Inc. 

California  Chassis  Co.,  Inc. 

Cardwell  Condenser  Co.,  Inc. 

Centralab,  Inc. 

Comell-Dubilier  Electric 
Co.,  Inc. 

Cowan  Publishing  Corp. 

International  Business 
Machines  Co.,  Inc. 

Marion  Electrical 

Instrument  Co.,  Inc, 

Miller  Coil  Co.,  Inc. 

Raypar,  Inc. 


Raytheon  Mfg.  Co.,  Inc. 
Sarkes-Tarzian,  Inc. 
Sprague  Electric  Co, 

Triad  Transformer  Co. 

Bob  Adams,  W6AVA 
Frank  Clement,  W6KPC 
AI  Cline,  W6LGU 
Temple  Ehmsen,  W7VS 
Ted  Gillett,  W6HX 
Bill  Glaser,  W60KG 
Bill  Guimont.WGYMD 
Ted  Henry,  W6UOU 
Herbert  Johnson,  W7GRA 
James  Lee,  W6VAT 
Earl  Lucas,  W2JT 
Bill  Mauzey,  W6WWQ 
Ken  Pierce,  W6SLQ 
Don  Stoner,  W6TNS 
Bob  Thompson,  K6SSJ 
Karl  Trovinger,  W6KMK 
Bill  Vandermav,  W7DET 
Dick  West,  W6IUG 
Edward  Willis,  W6TS 
Joseph  Jasgur 
(photography) 

B.  A.  Ontiveros,  W6FFF 
(drafting) 

Del  Rairigh,  W6ZAT 


CHAPTER  ONE 


Introduction  to  Radio 


The  field  of  radio  is  a division  of  the  much 
larger  field  of  electronics.  Radio  itself  is  such 
a broad  study  that  it  is  still  further  broken 
down  into  a number  of  smaller  fields  of  which 
only  shortwave  or  high-frequency  radio  is  cov- 
ered in  this  book.  Specifically  the  field  of  com- 
munication on  frequencies  from  1.8  to  450  meg- 
acycles is  taken  as  the  subject  matter  for  this 
work. 

The  largest  group  of  persons  interested  in 
the  subject  of  high-frequency  communication  is 
the  more  than  150,000  radio  amateurs  located 
in  nearly  all  countries  of  the  world.  Strictly 
speaking,  a radio  amateur  is  anyone  interested 
in  radio  non-commercially,  but  the  term  is  ordi- 
narily applied  only  to  those  hobbyists  possess- 
ing transmitting  equipment  and  a license  from 
the  government. 

It  was  for  the  radio  amateur,  and  particu- 
larly for  the  serious  and  more  advanced  ama- 
teur, that  most  of  the  equipment  described  in 
this  book  was  developed.  However,  in  each 
equipment  group  simple  items  also  are  shown 
for  the  student  or  beginner.  The  design  prin- 
ciples behind  the  equipment  for  high-frequency 
radio  communication  are  of  course  the  same 
whether  the  equipment  is  to  be  used  for  com- 
mercial, military,  or  amateur  purposes,  the 
principal  differences  lying  in  construction 
practices,  and  in  the  tolerances  and  safety 
factors  placed  upon  components. 

With  the  increasing  complexity  of  high- fre- 
quency communication,  resulting  primarily  from 
increased  utilization  of  the  available  spec- 
trum, it  becomes  necessary  to  delve  more  deep- 
ly into  the  basic  principles  underlying  radio 
communication,  both  from  the  standpoint  of 
equipment  design  and  operation  and  from  the 
standpoint  of  signal  propagation.  Hence,  it  will 
be  found  that  this  edition  of  the  RADIO  HAND- 
BOOK has  been  devoted  in  greater  proportion 


to  the  teaching  of  the  principles  of  equipment 
design  and  signalpropagation.lt  is  in  response 
to  requests  from  schools  and  agencies  of  the 
Department  of  Defense,  in  addition  to  persist- 
ent requests  from  the  amateur  radio  fraternity, 
that  coverage  of  these  principles  has  been  ex- 
panded. 

1-1  Amateur  Radio 

Amateur  radio  is  a fascinating  hobby  with 
many  phases.  So  strong  is  the  fascination  of- 
fered by  this  hobby  that  many  executives,  en- 
gineers, and  military  and  commercial  operators 
enjoy  amateur  radio  as  an  avocation  even 
though  they  are  also  engaged  in  the  radio  field 
commercially.  It  captures  and  holds  the  inter- 
est of  many  people  in  all  walks  of  life,  and  in 
all  countries  of  the  world  where  amateur  acti- 
vities are  permitted  by  law. 

Amateurs  have  rendered  much  public  ser- 
vice through  furnishing  communications  to  and 
from  the  outside  world  in  cases  where  disaster 
has  isolated  an  area  by  severing  all  wire  com- 
munications. Amateurs  have  a proud  record  of 
heroism  and  service  in  such  occasion.  Many 
expeditions  to  remote  places  have  been  kept 
in  touch  with  home  by  communication  with  ama- 
teur stations  on  the  high  frequencies.  The  ama- 
teur’s fine  record  of  performance  with  the 
’’wireless’*  equipment  of  World  War  I has  been 
surpassed  by  his  outstanding  service  in  World 
War  II. 

By  the  time  peace  came  in  the  Pacific  in 
the  summer  of  1945,  many  thousand  amateur 
operators  were  serving  in  the  allied  armed 
forces.  They  had  supplied  the  army,  navy, 
marines,  coast  guard,  merchant  marine,  civil 
service,  war  plants,  and  civilian  defense  or- 
ganizations with  trained  personnel  for  radio, 


12  Introduction  to  Radio 


THE  RADIO 


radar,  wire,  and  visual  communications  and 
for  teaching.  Even  now,  at  the  time  of  this 
writing,  amateurs  are  being  called  back  into 
the  expanded  defense  forces,  are  returning  to 
defense  plants  where  their  skills  are  critically 
needed,  and  are  being  organized  into  communi- 
cation units  as  an  adjunct  to  civil  defense 
groups. 

1-2  Station  and  Operator  Licenses 

Every  radio  transmitting  station  in  the 
United  States  no  matter  how  low  its  power 
must  have  a license  from  the  federal  govern- 
ment before  being  operated;  some  classes  of 
stations  must  have  a permit  from  the  govern- 
ment even  before  being  constructed.  And  every 
operator  of  a transmitting  station  must  have 
an  operator’s  license  before  operating  a trans- 
mitter. There  are  no  exceptions.  Similar  laws 
apply  in  practically  every  major  country. 

'Classes  of  Amateur  There  are  at  present  six 
Operator  Licenses  classes  of  amateur  oper- 
ator licenses  which  have 
been  authorized  by  the  Federal  Communica- 
tions Commission.  These  classes  differ  in 
many  respects,  so  each  will  be  discussed 
briefly. 

(a)  Amateur  Extra  Class.  This  class  of  li- 
cense is  available  to  any  U.  S.  citizen  who  at 
any  time  has  held  for  a period  of  two  years  or 
more  a valid  amateur  license,  issued  by  the 
FCC,  excluding  licenses  of  the  Novice  and 
Technician  Classes.  The  examination  for  the 
license  includes  a code  test  at  20  words  per 
minute,  the  usual  tests  covering  basic  amateur 
practice  and  general  amateur  regulations,  and 
an  additional  test  on  advanced  amateur  prac- 
tice. All  amateur  privileges  ate  accorded  the 
holders  of  this  operator’s  license. 

(b)  General  Class.  This  class  of  amateur 
license  is  equivalent  to  the  old  Amateur  Class 
B license,  and  accords  to  the  holders  all  ama- 
teur privileges  except  those  which  may  be  set 
aside  for  holders  of  the  Amateur  Extra  Class 
license.  This  class  of  amateur  operator’s  li- 
cense is  available  to  any  U.  S.  citizen.  The 
examination  for  the  license  includes  a code 
test  at  13  words  per  minute,  and  the  usual  ex- 
aminations coveting  basic  amateur  practice 
and  general  amateur  regulations. 

(c)  Conditional  Class.  This  class  of  ama- 
teur license  and  the  privileges  accorded  by  it 
ate  equivalent  to  the  General  Class  license. 
However,  the  license  can  be  issued  only  to 
those  whose  residence  is  more  than  125  miles 
airline  from  the  neatest  location  at  which  FCC 
examinations  are  held  at  intervals  of  not  more 
than  three  months  for  the  General  Class  ama- 
teur operator  license,  or  to  those  who  for  any 


of  several  specified  reasons  are  unable  to  ap- 
pear for  examination. 

(d)  Technician  Class.  This  is  a new  class 
of  license  which  is  available  to  any  citizen  of 
the  United  States.  The  examination  is  the  same 
as  that  for  the  General  Class  license,  except 
that  the  code  test  is  at  a speed  of  5 words  per 
minute.  The  holder  of  a Technician  class  li- 
cense is  accorded  all  authorized  amateur  privi- 
leges in  the  amateur  frequency  bands  above 
220  megacycles,  and  in  the  50-Mc.  band. 

(e)  Novice  Class.  This  is  a new  class  of 
license  which  is  available  to  any  U.  S.  citizen 
who  has  not  previously  held  an  amateur  li- 
cense of  any  class  issued  by  any  agency  of 
the  U.  S.  government,  military  or  civilian.  The 
examination  consists  of  a code  test  at  a speed 
of  5 words  per  minute,  plus  an  examination  on 
the  rules  and  regulations  essential  to  begin- 
ner’s operation,  including  sufficient  elemen- 
tary radio  theory  for  the  understanding  of  those 
rules.  The  Novice  Class  of  license  affords 
severely  restricted  privileges,  is  valid  for  only 
a period  of  one  year  (as  contrasted  to  all  other 
classes  of  amateur  licenses  which  run  for  a 
term  of  five  years),  and  is  not  renewable. 

All  Novice  and  Technician  class  examina- 
tions are  given  by  volunteer  examiners,  as  reg- 
ular examinations  for  these  two  classes  are 
not  given  in  FCC  offices.  Amateur  radio  clubs 
in  the  larger  cities  have  established  examin 
ing  committees  to  assist  would-be  amateurs 
of  the  area  in  obtaining  their  Novice  and  Tech- 
nician licenses. 

1-3  The  Amateur  Bands 

Certain  small  segments  of  the  radio  frequen- 
cy spectrum  between  1500  kc.  and  10,000  Me. 
ate  reserved  for  operation  of  amateur  radio 
stations.  These  segments  ate  in  general  agree- 
ment throughout  the  world,  although  certain 
parts  of  different  amateur  bands  may  be  used 
for  other  purposes  in  various  geographic  re- 
gions. In  particular,  the  40-meter  amateur  band 
is  used  legally  (and  illegally)  for  short  wave 
broadcasting  by  many  countries  in  Europe, 
Africa  and  Asia.  Parts  of  the  80-meter  band 
are  used  for  short  distance  marine  work  in  Eu- 
rope, and  for  broadcasting  in  South  America. 
The  amateur  bands  available  to  American  ra- 
dio amateurs  are: 

160  Meters  The  l60-meter  band  is  di- 

(1800  KC.-2000  Kc.)  vided  into  25-kilocycle 
segments  on  a regional 
basis,  with  day  and  night  power  limitations, 
and  is  available  for  amateur  use  provided  no 
interference  is  caused  to  the  Loran  (Long 
Range  Navigation)  stations  operating  in  this 
band.  This  band  is  least  affected  by  the  11- 


HANDBOOK 


Amateur  Bands  13 


year  solar  sunspot  cycle.  The  Maximum  Us- 
able Frequency  (MUF)  even  during  the  years 
of  decreased  sunspot  activity  does  not  usually 
drop  below  4 Me.,  therefore  this  band  is  not 
subject  to  the  violent  fluctuations  found  on 
the  higher  frequency  bands.  DX  contacts  on 
on  this  band  are  limited  by  the  ionospheric 
absorption  of  radio  signals,  which  is  quite 
high.  During  winter  nighttime  hours  the  ab- 
sorption is  often  of  a low  enough  value  to  per- 
mit trans-oceanic  contacts  on  this  band.  On 
rare  occasions,  contacts  up  to  10,000  miles 
have  been  made.  As  a usual  rule,  however, 
160-meter  amateur  operation  is  confined  to 
ground-wave  contacts  or  single-skip  contacts 
of  1000  miles  or  less.  Popular  before  World 
War  II,  the  160-meter  band  is  now  only  sparse- 
ly occupied  since  many  areas  of  the  country 
are  blanketed  by  the  megawatt  pulses  of  the 
Loran  chains. 

80  Meters  The  80-meter  band  is  the 

(3500  Kc.-4000  Kc.)  most  popular  amateur 
band  in  the  continental 
United  States  for  local  ^’rag-chewing”  and 
traffic  nets.  During  the  years  of  minimum  sun- 
spot activity  the  ionospheric  absorption  on 
this  band  may  be  quite  low,  and  long  distance 
DX  contacts  are  possible  during  the  winter 
night  hours.  Daytime  operation,  in  general,  is 
limited  to  contacts  of  500  miles  or  less.  Dur- 
ing the  summer  months,  local  static  and  high 
ionospheric  absorption  limit  long  distance  con- 
tacts on  this  band.  As  the  sunspot  cycle  ad- 
vances and  the  MUF  rises,  increased  iono- 
spheric absorption  will  tend  to  degrade  the 
long  distance  possibilities  of  this  band.  At 
the  peak  of  the  sunspot  cycle,  the  80-meter 
band  becomes  useful  only  for  short-haul  com- 
munication. 

40  Meters  The  40-meter  band  is  high 

(7000  Kc. -7300  Kc.)  enough  in  frequency  to  be 
severely  affected  by  the 
11-year  sunspot  cycle.  During  years  of  mini- 
mum solar  activity,  the  MUF  may  drop  below 
7 Me.,  and  the  band  will  become  very  erratic, 
with  signals  dropping  completely  out  during 
the  night  hours.  Ionospheric  absorption  of  sig- 
nals is  not  as  large  a problem  on  this  band  as 
it  is  on  80  and  160  meters.  As  the  MUF  grad- 
ually rises,  the  skip-distance  will  increase  on 
40  meters , especially  during  the  winter  months. 
At  the  peak  of  the  solar  cycle,  the  daylight 
skip  distance  on  40  meters  will  be  quite  long, 
and  stations  within  a distance  of  500  miles  or 
so  of  each  other  will  not  be  able  to  hold  com- 
munication. DX  operation  on  the  40-meter  band 
is  considerably  hampered  by  broadcasting  sta- 
tions, propaganda  stations,  and  jamming  trans- 


mitters. In  Europe  and  Asia  the  band  is  in  a 
chaotic  state,  and  amateur  operation  in  this  re- 
gion is  severely  hampered. 

20  Meters  At  the  present  time, 

(14,000  Kc.-14,350  Kc.)  the  20-meter  band  is 
by  far  the  most  popular 
band  for  long  distance  contacts.  High  enough 
in  frequency  to  be  almost  obliterated  at  the 
bottom  of  the  solar  cycle,  the  band  neverthe- 
less provides  good  DX  contacts  during  years 
of  minimal  sunspot  activity.  At  the  present 
time,  the  band  is  open  to  almost  all  parts  of 
the  world  at  some  time  during  the  year.  Dur- 
ing the  summer  months,  the  band  is  active  un- 
til the  late  evening  hours,  but  during  the  win- 
ter months  the  band  is  only  good  for  a few 
hours  during  daylight.  Extreme  DX  contacts 
are  usually  erratic,  but  the  20-meter  band  is 
the  only  band  available  for  DX  operation  the 
year  around  during  the  bottom  of  the  DX  cycle. 
As  the  sunspot  count  increases  and  the  MUF 
rises,  the  20-meter  band  will  become  open  for 
longer  hours  during  the  winter.  The  maximum 
skip  distance  increases,  and  DX  contacts  are 
possible  over  paths  other  than  the  Great  Circle 
route.  Signals  can  be  heard  the  'Mong  paths,” 
180  degrees  opposite  to  the  Great  Circle  path. 
During  daylight  hours,  absorption  may  become 
apparent  on  the  20-meter  band,  and  all  signals 
except  very  short  skip  may  disappear.  On  the 
other  hand,  the  band  will  be  open  for  world- 
wide DX  contacts  all  night  long.  The  20-meter 
band  is  very  susceptible  to  ”fade-outs” 
caused  by  solar  disturbances,  and  all  except 
local  signals  may  completely  disappear  for 
periods  of  a few  hours  to  a day  or  so. 


15  Meters  This  is  a relatively 

(21,000  Kc.-21,450  Kc.)  new  band  for  radio 
amateurs  since  it  has 
only  been  available  for  amateur  operation 
since  1952.  Not  too  much  is  known  about  the 
characteristics  of  this  band,  since  it  has  not 
been  occupied  for  a full  cycle  of  solar  activi- 
ty. However,  it  is  reasonable  to  assume  that 
it  will  have  characteristics  similar  to  both  the 
20  and  10-meter  amateur  bands.  It  should  have 
a longer  skip  distance  than  20  meters  for  a 
given  time,  and  sporadic-E  (short-skip)  should 
be  apparent  during  the  winter  months.  During 
a period  of  low  sunspot  activity,  the  MUF  will 
rarely  rise  as  high  as  15  meters,  so  this  band 
will  be  ’’dead”  for  a large  part  of  the  year. 
During  the  next  few  years,  15-meter  activity 
should  pick  up  rapidly,  and  the  band  should 

support  extremely  long  DX  contacts.  Activity 
on  the  15-meter  band  is  limited  in  some  areas. 


14  Introduction  to  Radio 


THE  RADIO 


since  the  older  model  TV  receivers  have  a 
21  Me.  i-f  channel,  which  falls  directly  in  the 
15-nieter  band.  The  interference  problems 
brought  about  by  such  an  unwise  choice  of 
intermediate  frequency  often  restrict  operation 
on  this  band  by  amateur  stations  unfortunate 
enough  to  be  situated  near  such  an  obsolete 
receiver. 

10-  Meters  During  the  peak  of  the 

(28,000  Kc.-29,700  Kc.)  sunspot  cycle,  the  IO- 
meter band  is  without 
doubt  the  most  popular 
amateur  band.  The  combination  of  long  skip 
and  low  ionospheric  absorption  make  reliable 
DX  contacts  with  low  powered  equipment  pos- 
sible. The  great  width  of  the  band  (1700  kc.) 
provides  room  for  a large  number  of  amateurs. 
The  long  skip  (1500  miles  or  so)  prevents  near- 
by amateurs  from  hearing  each  other,  thus 
dropping  the  interference  level.  During  the  win- 
ter months,  sporadic-E  (short  skip)  signals 
up  to  1200  miles  or  so  will  be  heard.  The  IO- 
meter band  is  poorest  in  the  summer  months, 
even  during  a sunspot  maximum.  Extremely 
long  daylight  skip  is  common  on  this  band,  and 
and  in  years  of  high  MUF  the  10-metet  band 
will  support  intercontinental  DX  contacts  dur- 
ing daylight  hours. 

The  second  harmonic  of  stations  operating 
in  the  10-metet  band  falls  directly  into  tele- 
vision channel  2,  and  the  higher  harmonics  of 
10-metet  transmitters  fall  into  the  higher  TV 
channels.  This  harmonic  problem  seriously 
curtailed  amateur  10-meter  operation  during 
the  late  40’s.  However,  with  the  new  circuit 
techniques  and  TVI  precautionary  measures 
stressed  in  this  Handbook,  10-metet  operation 
should  cause  little  or  no  interference  to  near- 
by television  receivers  of  modern  design. 

Six  Meters  At  the  peak  of  the  sunspot 

(50  Me. -54  Me.)  cycle,  the  MUF  occasional- 
ly rises  high  enough  to  per- 
mit DX  contacts  up  to  10,000  miles  or  so  on 
6 meters.  Activity  on  this  band  during  such  a 
period  is  often  quite  high.  Interest  in  this  band 
wanes  during  a period  of  lesser  solar  activity, 
as  contacts,  as  a rule,  are  restricted  to  short- 
skip  work.  The  proximity  of  the  6-meter  band 
to  television  channel  2 often  causes  interfer- 
ence problems  to  amateurs  located  in  areas 
where  channel  2 is  active.  As  the  sunspot  cy- 
cle increases,  activity  on  the  6-metet  band  will 
increase. 

The  V-H-F  Bands  The  v-h-f  bands  are 

(Two  Meters  and  “Up”)  the  least  affected  by 
the  vagaries  of  the 
sunspot  cycle  and  the  Heaviside  layer.  Their 
predominant  use  is  for  reliable  communication 
over  distances  of  150  miles  or  less.  These 


bands  are  sparsely  occupied  in  the  rural  sec- 
tions of  the  United  States,  but  are  quite  heavi- 
ly congested  in  the  urban  areas  of  high  popu- 
lation. 

In  recent  years  it  has  been  found  that  v-h-f 
signals  are  propagated  by  other  means  than  by 
line-of-sight  transmission.  "Scatter  signals,” 
Aurora  reflection,  and  air-mass  boundary  bend- 
ing are  responsible  for  v-h-f  communication  up 
to  1200  miles  or  so.  Weather  conditions  will 
often  affect  long  distance  communication  on 
the  2-metet  band,  and  all  the  v-h-f  bands  are 
particularly  sensitive  to  this  condition. 

The  other  v-h-f  bands  have  had  insufficient 
occupancy  to  provide  a clear  picture  of  their 
characteristics.  In  general,  they  behave  much 
as  does  the  2-meter  band,  with  the  weather 
effects  becoming  more  pronounced  on  the  high- 
er frequency  bands. 

1-4  Starting  Your  Study 

When  you  start  to  prepare  yourself  for  the 
amateur  examination  you  will  find  that  the  cir- 
cuit diagrams,  tube  characteristic  curves,  and 
formulas  appear  confusing  and  difficult  of  un- 
derstanding. But  after  a few  study  sessions 
one  becomes  sufficiently  familiar  with  the 
notation  of  the  diagrams  and  the  basic  con- 
cepts of  theory  and  operation  so  that  the  ac- 
quisition of  further  knowledge  becomes  easier 
and  even  fascinating. 

As  it  takes  a considerable  time  to  become 
proficient  in  sending  and  receiving  code,  it  is 
a good  idea  to  intersperse  technical  study  ses- 
sions with  periods  of  code  practice.  Many 
short  code  practice  sessions  benefit  one  more 
than  a small  number  of  longer  sessions.  Alter- 
nating between  one  study  and  the  other  keeps 
the  student  from  getting  "stale”  since  each 
type  of  study  serves  as  a sort  of  respite  from 
the  other. 

When  you  have  practiced  the  code  long 
enough  you  will  be  able  to  follow  the  gist  of 
the  slower  sending  stations.  Many  stations 
send  very  slowly  when  working  other  stations 
at  great  distances.  Stations  repeat  their  calls 
many  times  when  calling  other  stations  before 
contact  is  established,  and  one  need  not  have 
achieved  much  code  proficiency  to  make  out 
their  calls  and  thus  determine  their  location. 

The  Code  The  applicant  for  any  class  of  ama- 
teur operator  license  must  be  able 
to  send  and  receive  the  Continental  Code 
(sometimes  called  the  International  Morse 
Code).  The  speed  requited  for  the  sending  and 
receiving  test  may  be  either  5,  13,  or  20  words 
per  minute,  depending  upon  the  class  of  li- 
cense, assuming  an  average  of  five  characters 
to  the  word  in  each  case.  The  sending  and  re- 


HANDBOOK 


Learning  the  Code  15 


A 


N 


1 


B *••• 

D 

E • 

F •••• 

H •••• 

I •• 


3 — 

4 .•••-■ 


R • — • 
S ••• 

u 

v •••« 


5 

6 

7 

8 
9 
0 


K 

L 

M 


0 MEANS  ZERO.  AND  IS  WRITTEN  IN  THIS 

WAY  TO  DISTINGUISH  IT  FROM  THE  LETTER  *0". 
IT  OFTEN  IS  TRANSMITTED  INSTEAD  AS  ONE 
LONG  DASH  (EQUIVALENT  TO  5 DOTS) 


PERIOD  (.) 

COMMA  (,) 

INTERROGATION  (7) 

QUOTATION  MARK  (”  ) ••■•••• 

COLON  (0  Mas*** 

SEMICOLON  (i) 

PARENTHESIS  C)  aaaasaa 


WAITSIGNCASJ  ••••• 

DOUBLE  DASH  (BREAK)  ••••aa 

ERROR  (ERASE  SIGN)  •••••••• 

FRACTION  BAR  (/>  ••••• 

END  OF  MESSAGE  (AR) 

END  OF  TRANSMISSION  (SK)  •••••• 

INTERNAT.  DISTRESS  SIG,  (SOS)  ••••aa»»»» 


Figure  1 

The  Continental  (or  International  Morse)  Code  is  used  for  substantially  all  non^automatic  radio 
communication.  DO  NOT  memorise  from  the  printed  page;  code  Is  a language  of  SOUND,  and 
must  not  be  learned  visually;  learn  by  listening  as  explained  In  the  text. 


ceiving  tests  tun  for  five  minutes,  and  one 
minute  of  errorless  transmission  or  reception 
must  be  accomplished  within  the  five-minute 
interval. 

If  the  code  test  is  failed,  the  applicant  must 
wait  at  least  one  month  before  he  may  again 
appear  for  another  test.  Approximately  30%  of 
amateur  applicants  fail  to  pass  the  test.  It 
should  be  expected  that  nervousness  and  ex- 
citement will  at  least  to  some  degree  tempo- 
rarily lower  the  applicant’s  code  ability.  The 
best  prevention  against  this  is  to  master  the 
code  at  a little  greater  than  the  required  speed 
under  ordinary  conditions.  Then  if  you  slow 
down  a little  due  to  nervousness  during  a test 
the  result  will  not  prove  fatal. 

Memorizing  There  is  no  shortcut  to  code  pro- 
the  Code  ficiency.  To  memorize  the  al- 
phabet entails  but  a few  eve- 
nings of  diligent  application,  but  considerable 
time  is  required  to  build  up  speed.  The  exact 
time  required  depends  upon  the  individual’s 
ability  and  the  regularity  of  practice. 

While  the  speed  of  learning  will  naturally 
vary  greatly  with  different  individuals,  about 
70  hours  of  practice  (no  practice  period  to  be 
over  30  minutes)  will  usually  suffice  to  bring 
a speed  of  about  13  w.p.m.;  16  w.p.m.  requires 
about  120  hours;  20  w.p.m.,  175  hours. 


Since  code  reading  requires  that  individual 
letters  be  recognized  instantly,  any  memoriz- 
ing scheme  which  depends  upon  orderly  se- 
quence, such  as  learning  all  ”dah“  letters 
and  all  ”dit"  letters  in  separate  groups,  is  to 
be  discouraged.  Before  beginning  with  a code 
practice  set  it  is  necessary  to  memorize  the 
whole  alphabet  perfectly.  A good  plan  is  to 
study  only  two  or  three  letters  a day  and  to 
drill  with  those  letters  until  they  become  part 
of  your  consciousness.  Mentally  translate  each 
day’s  letters  into  their  sound  equivalent 
wherever  they  are  seen,  on  signs,  in  papers, 
indoors  and  outdoors.  Tackle  two  additional 
letters  in  the  code  cheirt  each  day,  at  the  same 
time  reviewing  the  characters  already  learned. 

Avoid  memorizing  by  routine.  Be  able  to 
sound  out  any  letter  immediately  without  so 
much  as  hesitating  to  think  about  the  letters 
preceding  or  following  the  one  in  question. 
Know  C,  for  example,  apart  from  the  sequence 
ABC.  Skip  about  among  all  the  characters 
learned,  and  before  very  long  sufficient  letters 
will  have  been  acquired  to  enable  you  to  spell 
out  simple  words  to  yourself  in  ”dit  dahs." 
This  is  interesting  exercise,  and  for  that  rea- 
son it  is  good  to  memorize  all  the  vowels  first 
and  the  most  common  consonants  next. 

Actual  code  practice  should  start  only  when 
the  entire  alphabet,  the  numerals,  period,  com- 


16  Introduction  to  Radio 


the  radio 


a«.a  •••••«■» 
e • • • • 

Figure  2 

These  code  characters  are  used  in  languages 
other  than  English.  They  may  occasionally 
be  encountered  so  it  is  well  to  know  them. 


ma,  and  question  mark  have  been  memorized 
so  thoroughly  that  any  one  can  be  sounded 
without  the  slightest  hesitation.  Do  not  bother 
with  other  punctuation  or  miscellaneous  sig- 
nals until  later. 

Sound  — Each  letter  and  figure  must  be 
No*  Sight  memorized  by  its  sound  rather 
than  its  appearance.  Code  is  a 
system  of  sound  communication,  the  same  as 
is  the  spoken  word.  The  letter  A,  for  example, 
is  one  short  and  one  long  sound  in  combina- 
tion sounding  like  dit  dah,  and  it  must  be  re- 
membered as  such,  and  not  as  "dot  dash.” 

Practice  Time,  patience,  and  regularity  are 
required  to  learn  the  code  properly. 
Do  not  expect  to  accomplish  it  within  a few 
days. 

Don’t  practice  too  long  at  one  stretch;  it 
does  more  harm  than  good.  Thirty  minutes  at 
a time  should  be  the  limit. 

Lack  of  regularity  in  practice  is  the  most 
common  cause  of  lack  of  progress.  Irregular 
practice  is  very  little  better  than  no  practice 
at  all.  Write  down  what  you  have  heard;  then 
forget  it;  do  not  look  hack.  If  your  mind  dwells 
even  for  an  instant  on  a signal  about  which 
you  have  doubt,  you  will  miss  the  next  few 
characters  while  your  attention  is  diverted. 

While  various  automatic  code  machines, 
phonograph  records,  etc.,  will  give  you  prac- 
tice, by  far  the  best  practice  is  to  obtain  a 
study  companion  who  is  also  interested  in 
learning  the  code.  When  you  have  both  memo- 
rized the  alphabet  you  can  start  sending  to 
each  other.  Practice  with  a key  and  oscillator 
or  key  and  buzzer  generally  proves  superior 
to  all  automatic  equipment.  Two  such  sets 
operated  between  two  rooms  are  fine— or  be- 
tween your  house  and  his  will  be  just  that 
much  better.  Avoid  talking  to  your  partner 
while  practicing.  If  you  must  ask  him  a ques- 


tion, do  it  in  code.  It  makes  more  interesting 
practice  than  confining  yourself  to  random 
practice  material. 

When  two  co-learners  have  memorized  the 
code  and  ate  ready  to  start  sending  to  each 
other  for  practice,  it  is  a good  idea  to  enlist 
the  aid  of  an  experienced  operator  for  the  first 
practice  session  or  two  so  that  they  will  get 
an  idea  of  how  properly  formed  characters 
sound. 

During  the  first  practice  period  the  speed 
should  be  such  that  substantially  solid  copy 
can  be  made  without  strain.  Never  mind  if  this 
is  only  two  or  three  words  per  minute.  In  the 
next  period  the  speed  should  be  increased 
slightly  to  a point  where  nearly  all  of  the 
characters  can  be  caught  only  through  con- 
scious effort.  When  the  student  becomes  pro- 
ficient at  this  new  speed,  another  slight  in- 
crease may  be  made,  progressing  in  this  man- 
ner until  a speed  of  about  16  words  per  minute 
is  attained  if  the  object  is  to  pass  the  amateur 
13-word  per  minute  code  test.  The  margin  of 
3 w.p.m.  is  recommended  to  overcome  a possi- 
ble excitement  factor  at  examination  time. 
Then  when  you  take  the  test  you  don’t  have  to 
worry  about  the  "jitters”  or  an  "off  day.” 

Speed  should  not  be  increased  to  a new 
level  until  the  student  finally  makes  solid 
copy  with  ease  for  at  least  a five-minute 
period  at  the  old  level.  How  frequently  in- 
creases of  speed  can  be  made  depends  upon 
individual  ability  and  the  amount  of  practice. 
Each  increase  is  ^t  to  prove  disconcerting, 
but  remember  "you  are  never  learning  when 
you  are  comfortable.” 

A number  of  amateurs  are  sending  code 
practice  on  the  air  on  schedule  once  or  twice 
each  week;  excellent  practice  can  be  obtained 
after  you  have  bought  or  constructed  your  re- 
ceiver by  taking  advantage  of  these  sessions. 

If  you  live  in  a medium-size  or  large  city, 
the  chances  are  that  there  is  an  amateur  radio 
club  in  your  vicinity  which  offers  free  code 
practice  lessons  periodically. 

Skill  When  you  listen  to  someone  speaking 
you  do  not  consciously  think  how  his 
words  are  spelled.  This  is  also  true  when  you 
read.  In  code  you  must  train  your  ears  to  read 
code  just  as  your  eyes  were  trained  in  school 
to  read  printed  matter.  With  enough  practice 
you  acquire  skill,  and  from  skill,  speed.  In 
other  words,  it  becomes  a habit,  something 
which  can  be  done  without  conscious  effort. 
Conscious  effort  is  fatal  to  speed;  we  can’t 
think  rapidly  enough;  a speed  of  25  words  a 
minute,  which  is  a common  one  in  commercial 
operations,  means  125  characters  per  minute 
or  more  than  two  pet  second,  which  leaves 
no  time  for  conscious  thinking. 


HANDBOOK 


Learning  the  Code  17 


Perfect  Formation  When  transmitting  on  the 

of  Characters  code  practice  set  to  your 

partner,  concentrate  on  the 
quality  of  your  sending,  not  on  your  speed. 
Your  partner  will  appreciate  it  and  he  could 
not  copy  you  if  you  speeded  up  anyhow. 

If  you  want  to  get  a reputation  as  having  an 
excellent  **fist**  on  the  air,  just  remember  that 
speed  alone  won^t  do  the  trick.  Proper  execu- 
tion  of  your  letters  and  spacing  will  make 
much  more  of  an  impression.  Fortunately,  as 
you  get  so  that  you  can  send  evenly  and  accu- 
rately, your  sending  speed  will  automatically 
increase.  Remember  to  try  to  see  how  evenly 
you  can  send,  and  how  fast  you  can  receive. 
Concentrate  on  making  signals  properly  with 
your  key.  Perfect  formation  of  characters  is 
paramount  to  everything  else.  Make  every  sig- 
nal right  no  matter  if  you  have  to  practice  it 
hundreds  or  thousands  of  times.  Never  allow 
yourself  to  vary  the  slightest  from  perfect  for- 
mation once  you  have  learned  it. 

If  possible,  get  a good  operator  to  listen  to 
your  sending  for  a short  time,  asking  him  to 
criticize  even  the  slightest  imperfections. 

Timing  It  is  of  the  utmost  importance  to 
maintain  uniform  spacing  in  charac- 
ters and  combinations  of  characters.  Lack  of 
uniformity  at  this  point  probably  causes  be- 
ginners more  trouble  than  any  other  single  fac- 
tor. Every  dot,  every  dash,  and  every  space 
must  be  correctly  timed.  In  other  words,  ac- 
curate timing  is  absolutely  essential  to  intel- 
ligibility, and  timing  of  the  spaces  between 
the  dots  and  dashes  is  just  as  important  as 
the  lengths  of  the  dots  and  dashes  themselves. 

The  characters  are  timed  with  the  dot  as  a 
’^yardstick. A standard  dash  is  three  times 
as  long  as  a dot.  The  spacing  between  parts 
of  the  same  letter  is  equal  to  one  dot;  the 
space  between  letters  is  equal  to  three  dots, 
and  that  between  words  equal  to  five  dots. 

The  rule  for  spacing  between  letters  and 
words  is  not  strictly  observed  when  sending 
slower  than  about  10  words  per  minute  for  the 
benefit  of  someone  learning  the  code  and  de- 
siring receiving  practice.  When  sending  at, 
say,  5 w.p.m.,  the  individual  letters  should  be 
made  the  same  as  if  the  sending  rate  were 
about  10  w.p.m.,  except  that  the  spacing  be- 
tween letters  and  words  is  greatly  exaggerated. 
The  reason  for  this  is  obvious.  The  letter  L, 
for  instance,  will  then  sound  exactly  the  same 
at  10  w.p.m.  as  at  5 w.p.m.,  and  when  the 
speed  is  increased  above  5 w.p.m.  the  student 
will  not  have  to  become  familiar  with  what 
may  seem  to  him  like  a new  sound,  although 
it  is  in  reality  only  a faster  combination  of 
dots  and  dashes.  At  the  greater  speed  he  will 
merely  have  to  learn  the  identification  of  the 
same  sound  without  taking  as  long  to  do  so. 


pocbplop^, 

'imm  « m 

1 

1 

B 

c 

icjcixxpoqcxDccoccd^^ 

‘‘mi  w 

1— 

< 

o'  N E 

Figure  3 

Diagram  illustrating  relative  lengths  of 
dashes  and  spaces  referred  to  the  duration 
of  o dot.  A dash  is  exactly  equal  in  duration 
to  three  dots;  spaces  between  parts  of  a 
letter  equal  one  dot;  those  between  letters, 
three  dots;  space  between  words,  five  dots. 
Note  that  a slight  increase  between  two  parts 
of  a letter  will  make  it  sound  like  two 
letters. 


Be  particularly  careful  of  letters  like  B. 
Many  beginners  seem  to  have  a tendency  to 
leave  a longer  space  after  the  dash  than  that 
which  they  place  between  succeeding  dots, 
thus  making  it  sound  like  TS.  Similarly,  make 
sure  that  you  do  not  leave  a longer  space  after 
the  first  dot  in  the  letter  C than  you  do  be- 
tween other  parts  of  the  same  letter;  otherwise 
it  will  sound  like  NN. 

Sending  vs.  Once  you  have  memorized  the 
Receiving  code  thoroughly  you  should  con- 
centrate on  increasing  your  re- 
ceiving  speed.  True,  if  you  have  to  practice 
with  another  newcomer  who  is  learning  the 
code  with  you,  you  will  both  have  to  do  some 
sending.  But  don’t  attempt  to  practice  sending 
just  for  the  sake  of  increasing  your  sending 
speed. 

When  transmitting  on  the  code  practice  set 
to  your  partner  so  that  he  can  get  receiving 
practice,  concentrate  on  the  quality  of  your 
sending,  not  on  your  speed. 

Because  it  is  comparatively  easy  to  learn 
to  send  rapidly,  especially  when  no  particular 
care  is  given  to  the  quality  of  sending,  many 
operators  who  have  just  received  their  licenses 
get  on  the  air  and  send  mediocre  or  worse  code 
at  20  w.p.m.  when  they  can  barely  receive 
good  code  at  13-  Most  oldtimers  remember  their 
own  period  of  initiation  and  are  only  too  glad 
to  be  patient  and  considerate  if  you  tell  them 
that  you  are  a newcomer.  But  the  surest  way 
to  incur  their  scorn  is  to  try  to  impress  them 
with  your  **lightning  speed,”  and  then  to  re- 
quest them  to  send  more  slowly  when  they 
come  back  at  you  at  the  same  speed. 

Stress  your  copying  ability;  never  stress 
your  sending  ability.  ^It  should  be  obvious  that 

that  if  you  try  to  send  faster  than  you  can  re- 
ceive, your  ear  will  not  recognize  any  mis- 
takes which  your  hand  may  make. 


18  Introduction  to  Radio 


THE  RADIO 


Figure  4 

PROPER  POSITION  OF  THE  FINGERS  FOR 
OPERATING  A TELEGRAPH  KEY 

The  fingers  hold  the  knob  and  act  as  a cush- 
ion. The  hand  rests  lightly  on  the  key.  The 
muscles  of  the  forearm  provide  the  power, 
the  wrist  acting  as  the  fulcrum.  The  power 
should  not  come  from  the  fingers,  but  rather 
from  the  forearm  muscles. 


Using  the  Key  Figure  4 shows  the  proper  posi- 
tion of  the  hand,  fingers  and 
wrist  when  manipulating  a telegraph  or  radio 
key.  The  forearm  should  rest  naturally  on  the 
desk.  It  is  preferable  that  the  key  be  placed 
far  enough  back  from  the  edge  of  the  table 
(about  18  inches)  that  the  elbow  can  rest  on 
the  table.  Otherwise,  pressure  of  the  table 
edge  on  the  arm  will  tend  to  hinder  the  circu- 
lation of  the  blood  and  weaken  the  ulnar  nerve 
at  a point  where  it  is  close  to  the  surface, 
which  in  turn  will  tend  to  increase  fatigue 
considerably. 

The  knob  of  the  key  is  grasped  lightly  with 
the  thumb  along  the  edge;  the  index  and  third 
fingers  rest  on  the  top  towards  the  front  or  far 
edge.  The  hand  moves  with  a free  up  and  down 
motion,  the  wrist  acting  as  a fulcrum.  The 
power  must  come  entirely  from  the  arm  mus- 
cles. The  third  and  index  fingers  will  bend 
slightly  during  the  sending  but  not  because  of 
deliberate  effort  to  manipulate  the  finger  mus- 
cles. Keep  your  finger  muscles  just  tight 
enough  to  act  as  a cushion  for  the  arm  motion 
and  let  the  slight  movement  of  the  fingers  take 
care  of  itself.  The  key’s  spring  is  adjusted  to 
the  individual  wrist  and  should  be  neither  too 
stiff  nor  too  loose.  Use  a moderately  stiff  ten- 
sion at  first  and  gradually  lighten  it  as  you 
become  more  proficient.  The  separation  be- 
tween the  contacts  must  be  the  proper  amount 
for  the  desired  speed,  being  somewhat  under 
1/ 16  inch  for  slow  speeds  and  slightly  closer 
together  (about  1/32  inch)  for  faster  speeds. 
Avoid  extremes  in  either  direction. 

Do  not  allow  the  muscles  of  arm,  wrist,  or 


fingers  to  become  tense.  Send  with  a full,  free 
arm  movement.  Avoid  like  the  plague  any  fin- 
ger motion  other  than  the  slight  cushioning 
effect  mentioned  above. 

Stick  to  the  regular  hand  key  for  learning 
code.  No  other  key  is  satisfactory  for  this  pur- 
pose. Not  until  you  have  thoroughly  mastered 
both  sending  and  receiving  at  the  maximum 
speed  in  which  you  ate  interested  should  you 
tackle  any  form  of  automatic  or  semi-automatic 
key  such  as  the  Vibroplex  ("bug”)  or  an  elec- 
tronic key. 

Difficulties  Should  you  experience  difficulty 
in  increasing  your  code  speed 
after  you  have  once  memorized  the  characters, 
there  is  no  reason  to  become  discouraged.  It 
is  more  difficult  for  some  people  to  learn  code 
than  for  others,  but  there  is  no  justification 
for  the  contention  sometimes  made  that  "some 
people  just  can’t  learn  the  code.”  It  is  not  a 
matter  of  intelligence;  so  don’t  feel  ashamed 
if  you  seem  to  experience  a little  more  than 
the  usual  difficulty  in  learning  code.  Your  re- 
action time  may  be  a little  slower  or  your  co- 
ordination not  so  good.  If  this  is  the  case, 
remember  you  can  still  learn  the  code.  You 
may  never  learn  to  send  and  receive  at  40 
w.p.m.,  but  you  can  learn  sufficient  speed  for 
all  non-commercial  purposes  and  even  for  most 
commercial  purposes  if  you  have  patience, 
and  refuse  to  be  discouraged  by  the  fact  that 
others  seem  to  pick  it  up  more  rapidly. 

When  the  sending  operator  is  sending  just 
a bit  too  fast  for  you  (the  best  speed  for  prac- 
tice), you  will  occasionally  miss  a signal  or  a 
small  group  of  them.  When  you  do,  leave  a 
blank  space;  do  not  spend  time  futilely  trying 
to  recall  it;  dismiss  it,  and  center  attention 
on  the  next  letter;  otherwise  you’ll  miss  more. 
Do  not  ask  the  sender  any  questions  until  the 
transmission  is  finished. 

To  prevent  guessing  and  get  equal  practice 
on  the  less  common  letters,  depart  occasional- 
ly from  plain  language  material  and  use  a jum- 
ble of  letters  in  which  the  usually  less  com- 
monly used  letters  predominate. 

As  mentioned  before,  many  students  put  a 
greater  space  after  the  dash  in  the  letter  B 
than  between  other  parts  of  the  same  letter  so 
it  sounds  like  TS.  C,  F,  Q,  V,  X,  Y and  Z 
often  give  similar  trouble.  Make  a list  of  words 
or  arbitrary  combinations  in  which  these  let- 
ters predominate  and  practice  them,  both  send- 
ing and  receiving  until  they  no  longer  give  you 
trouble.  Stop  everything  else  and  stick  at 
them.  So  long  as  they  give  you  trouble  you  are 
not  ready  for  anything  else. 

Follow  the  same  procedure  with  letters 
which  you  may  tend  to  confuse  such  as  P and 
L,  which  are  often  confused  by  beginners. 


HANDBOOK 


Learning  the  Code 


19 


Figure  5 

THE  SIMPLEST  CODE  PRACTICE 
SET  CONSISTS  OF  A KEY  AND  A 
BUZZER 

The  buzzer  is  adjusted  to  give  a 
steady,  high-pitched  whine.  If  de- 
sired, the  phones  may  be  omitted, 
in  which  case  the  buzzer  should  be 
mounted  firmly  on  a sounding  board. 
Crystal,  magnetic,  or  dynamic  ear- 
phones may  be  used.  Additional 
sets  of  phones  should  be  connected 
in  parallel,  not  in  series. 


Keep  at  it  until  you  always  get  them  right 
without  having  to  stop  even  an  instant  to  think 
about  it. 

If  you  do  not  instantly  recognize  the  sound 
of  any  character,  you  have  not  learned  it;  go 
back  and  practice  your  alphabet  further.  You 
should  never  have  to  omit  writing  down  every 
signal  you  hear  except  when  .the  transmission 
is  too  fast  for  you. 

Write  down  what  you  hear,  not  what  you 
think  it  should  be.  It  is  surprising  how  often 
the  word  which  you  guess  will  be  wrong. 

Copying  Behind  All  good  operators  copy  sev- 
eral words  behind,  that  is, 
while  one  word  is  being  received,  they  are 
writing  down  or  typing,  say,  the  fourth  or  fifth 
previous  word.  At  first  this  is  very  difficult, 
but  after  sufficient  practice  it  will  be  found 
actually  to  be  easier  than  copying  close  up. 
It  also  results  in  more  accurate  copy  and  en- 
ables the  receiving  operator  to  capitalize  and 


1 = COLLECTOR 


SIMPLE  TRANSISTOR  CODE 
PRACTICE  OSCILLATOR 
An  inexpensive  Raytheon  CK-722  transistor 
requires  only  a single  IVz-volt  flashlight 
battery  for  power.  The  inductance  of  the  ear- 
phone windings  forms  part  of  the  oscillatory 
circuit.  The  pitch  of  the  note  may  be  changed 
by  varying  the  value  ef  the  two  capacitors 
connected  across  the  earphones. 


punctuate  copy  as  he  goes  along.  It  is  not  rec- 
ommended that  the  beginner  attempt  to  do  this 
until  he  can  send  and  receive  accurately  and 
with  ease  at  a speed  of  at  least  12  words  a 
minute. 

It  requires  a considerable  amount  of  train- 
ing to  dissociate  the  action  of  the  subcon- 
scious mind  from  the  direction  of  the  conscious 
mind.  It  may  help  some  in  obtaining  this  train- 
ing to  write  down  two  columns  of  short  words. 
Spell  the  first  word  in  the  first  column  out  loud 
while  writing  down  the  first  word  in  the  second 
column.  At  first  this  will  be  a bit  awkward, 
but  you  will  rapidly  gain  facility  with  practice. 
Do  the  same  with  all  the  words,  and  then  re- 
verse columns. 

Next  try  speaking  aloud  the  words  in  the  one 
column  while  writing  those  in  the  other  column; 
then  reverse  columns. 

After  the  foregoing  can  be  done  easily,  try 
sending  with  your  key  the  words  in  one  col- 
umn while  spelling  those  in  the  other.  It  won’t 
be  easy  at  first,  but  it  is  well  worth  keeping 
after  if  you  intend  to  develop  any  real  code 
proficiency.  Do  not  attempt  to  catch  up.  There 
is  a natural  tendency  to  close  up  the  gap,  and 
you  must  train  yourself  to  overcome  this. 

Next  have  your  code  companion  send  you  a 
word  either  from  a list  or  from  straight  text; 
do  not  write  it  down  yet.  Now  have  him  send 
the  next  word;  after  receiving  this  second 
word,  write  down  the  first  word.  After  receiv- 
ing the  third  word,  write  the  second  word;  and 
so  on.  Never  mind  how  slowly  you  must  go, 
even  if  it  is  only  two  or  three  words  per  minute. 
Stay  behind. 

It  will  probably  take  quite  a number  of  prac- 
tice sessions  before  you  can  do  this  with  any 
facility.  After  it  is  relatively  easy,  then  tty 
staying  two  words  behind;  keep  this  up  until 
it  is  easy.  Then  try  three  words,  four  words, 
and  five  words.  The  mote  you  practice  keep- 
ing received  material  in  mind,  the  easier  it 
will  be  to  stay  behind.  It  will  be  found  easier 
at  first  to  copy  material  with  which  one  is 
fairly  familiar,  then  gradually  switch  to  less 
familiar  material. 


20  Introduction  to  Radio 


Automatic  Code  The  two  practice  sets  which 
Machines  are  described  in  this  chapter 

are  of  most  value  when  you 
have  someone  with  whom  to  practice.  Automa- 
tic code  machines  are  not  recommended  to  any- 
one who  can  possibly  obtain  a companion  with 
whom  to  practice,  someone  who  is  also  inter- 
ested in  learning  the  code.  If  you  are  unable 
to  enlist  a code  partner  and  have  to  practice 
by  yourself,  the  best  way  to  get  receiving 
practice  is  by  the  use  of  a tape  machine  (auto- 
matic code  sending  machine)  with  several 
practice  t^es.  Or  you  can  use  a set  of  phono- 
graph code  practice  records.  The  records  are 
of  use  only  if  you  have  a phonograph  whose 
turntable  speed  is  readily  adjustable.  The  tape 
machine  can  be  rented  by  the  month  for  a rea- 
sonable fee. 

Once  you  can  copy  about  10  w.p.m.  you  can 
also  get  receiving  practice  by  listening  to  slow 
sending  stations  on  your  receiver.  Many  ama- 
teur stations  send  slowly  particularly  when 
working  far  distant  stations.  When  receiving 
conditions  are  particularly  poor  many  commer- 
cial stations  also  send  slowly,  sometimes  re- 
peating every  word.  Until  you  can  copy  around 
10  w.p.m,  your  receiver  isn’t  much  use,  and 
either  another  operator  or  a machine  or  records 
are  necessary  for  getting  receiving  practice 
after  you  have  once  memorized  the  code. 


Code  Practice  If  you  don’t  feel  too  foolish 
Sets  doing  it,  you  can  secure  a 

measure  of  code  practice  with 


the  help  of  a partner  by  sending  ^'dit-dah” 
messages  to  each  other  while  riding  to  work, 
eating  lunch,  etc.  It  is  better,  however,  to  use 
a buzzer  or  code  practice  oscillator  in  con- 
junction with  a regular  telegraph  key. 

As  a good  key  may  be  considered  an  invest- 
ment it  is  wise  to  make  a well-made  key  your 
first  purchase.  Regardless  of  what  type  code 
practice  set  you  use,  you  will  need  a key,  and 
later  on  you  will  need  one  to  key  your  trans- 
mitter. If  you  get  a good  key  to  begin  with, 
you  won’t  have  to  buy  another  one  later. 

The  key  should  be  rugged  and  have  fairly 
heavy  contacts.  Not  only  will  the  key  stand 
up  better,  but  such  a key  will  contribute  to 
the  ’’heavy”  type  of  sending  so  desirable  for 
radio  work.  Morse  (telegraph)  operators  use 
a ’’light”  style  of  sending  and  can  send  some- 
what faster  when  using  this  light  touch.  But, 
in  radio  work  static  and  interference  are  often 
present,  and  a slightly  heavier  dot  is  desir- 
able. If  you  use  a husky  key,  you  will  tind 
yourself  automatically  sending  in  this  manner. 

To  generate  a tone  simulating  a code  signal 
as  heard  on  a receiver,  either  a mechanical 
buzzer  or  an  audio  oscillator  maybe  used.  Fig- 
ure 5 shows  a simple  code-practice  set  using 
a buzzer  which  may  be  used  directly  simply 
by  mounting  the  buzzer  on  a sounding  board, 
or  the  buzzer  may  be  used  to  feed  from  one  to 
four  pairs  of  conventional  high-impedance 
phones. 

An  example  of  the  audio-oscillator  type  of 
code-practice  set  is  illustrated  in  figures  6 
and  7.  An  inexpensive  Raytheon  CK-722  trans- 
istor is  used  in  place  of  the  more  expensive, 
power  consuming  vacuum  tube.  A single  ”pen- 
lite”  1^-volt  cell  powers  the  unit.  The  coils 
of  the  earphones  form  the  inductive  portion 
of  the  resonant  circuit.  ’Phones  having  an 
impedance  of  2000  ohms  or  higher  should  be 
used.  Surplus  type  R-14  earphones  also  work 
well  with  this  circuit. 


Figure  7 

The  circuit  of  Figure  6 is  used  in  this 
miniature  transistorized  code  Practice 
oscillator.  Components  are  mounted  in  a 
small  plastic  case.  The  tronsisfor  is 
attached  to  a three  term/no/  phenolic 
mounting  strip.  Sub-miniature  jacks  are 
used  for  the  key  and  phones  connections. 
A hearing  aid  earphone  may  also  be  used, 
as  shown.  The  phone  is  stored  in  the 
plastic  case  when  not  in  use. 


CHAPTER  TWO 


Direct  Current  Circuits 


All  naturally  occurring  matter  (excluding 
artifically  produced  radioactive  substances)  is 
made  up  of  92  fundamental  constituents  called 
elements.  These  elements  can  exist  either  in 
the  free  state  such  as  iron,  oxygen,  carbon, 
copper,  tungsten,  and  aluminum,  or  in  chemi- 
cal unions  commonly  called  compounds.  The 
smallest  unit  which  still  retains  all  the  origi- 
nal characteristics  of  an  element  is  the  atom. 

Combinations  of  atoms,  or  subdivisions  of 
compounds,  result  in  another  fundamental 
unit,  the  molecule.  The  molecule  is  the  small- 
est unit  of  any  compound.  All  reactive  ele- 
ments when  in  the  gaseous  state  also  exist 
in  the  molecular  form,  made  up  of  two  or  more 
atoms.  The  nonreactive  gaseous  elements 
helium,  neon,  argon,  krypton,  xenon,  and 
radon  are  the  only  gaseous  elements  that  ever 
exist  in  a stable  monatomic  state  at  ordinary 
temperatures. 

2-1  The  Atom 

An  atom  is  an  extremely  small  unit  of 
matter  — there  are  literally  billions  of  them 
making  up  so  small  a piece  of  material  as  a 
speck  of  dust.  To  understand  the  basic  theory 
of  electricity  and  hence  of  radio,  we  must  go 
further  and  divide  the  atom  into  its  main 
components,  a positively  charged  nucleus  and 
a cloud  of  negatively  charged  particles  that 
surround  the  nucleus.  These  particles,  swirling 
around  the  nucleus  in  elliptical  orbits  at  an 
incredible  rate  of  speed,  are  called  orbital 
electrons. 

It  is  upon  the  behavior  of  these  electrons 
when  freed  from  the  atom,  that  depends  the 
study  of  electricity  and  radio,  as  well  as 
allied  sciences.  Actually  it  is  possible  to  sub- 
divide the  nucleus  of  the  atom  into  a dozen  or 


so  different  particles,  but  this  further  sub- 
division can  be  left  to  quantum  mechanics  and 
atomic  physics.  As  far  as  the  study  of  elec- 
tronics is  concerned  it  is  only  necessary  for 
the  reader  to  think  of  the  normal  atom  as  being 
composed  of  a nucleus  having  a net  positive 
charge  that  is  exactly  neutralized  by  the  one 
or  more  orbital  electrons  surrounding  it. 

The  atoms  of  different  elements  differ  in 
respect  to  the  charge  on  the  positive  nucleus 
and  in  the  number  of  electrons  revolving 
around  this  charge.  They  range  all  the  way 
from  hydrogen,  having  a net  charge  of  one 
on  the  nucleus  and  one  orbital  electron,  to 
uranium  with  a net  charge  of  92  on  the  nucleus 
and  92  orbital  electrons.  The  number  of  orbital 
electrons  is  called  the  atomic  number  of  the 
element. 

Action  of  the  From  the  above  it  must  not  be 
Electrons  thought  that  the  electrons  re- 
volve in  a haphazard  manner 
around  the  nucleus.  Rather,  the  electrons  in 
an  element  having  a large  atomic  number  are 
grouped  into  rings  having  a definite  number  of 
electrons.  The  only  atoms  in  which  these  rings 
are  completely  filled  ate  those  of  the  inert 
gases  mentioned  before;  all  other  elements 
have  one  or  more  uncompleted  tings  of  elec- 
trons. If  the  uncompleted  ring  is  nearly  empty, 
the  element  is  metallic  in  character,  being 
most  metallic  when  there  is  only  one  electron 
in  the  outer  ring.  If  the  incomplete  ring  lacks 
only  one  or  two  electrons,  the  element  is 
usually  non-metallic.  Elements  with  a ring 
about  half  completed  will  exhibit  both  non- 
metallic  and  metallic  characteristics;  carbon, 
silicon,  germanium,  and  arsenic  are  examples. 
Such  elements  are  called  semi-conductors. 

In  metallic  elements  these  outer  ring  elec- 
trons are  rather  loosely  held.  Consequently, 


21 


22  Direct  Current  Circuits 


THE  RADIO 


there  is  a continuous  helter-skelter  movement 
of  these  electrons  and  a continual  shifting 
from  one  atom  to  another.  The  electrons  which 
move  about  in  a substance  are  called  free 
electrons,  and  it  is  the  ability  of  these  elec- 
trons to  drift  from  atom  to  atom  which  makes 
possible  the  electric  current. 

Conductors  and  If  the  free  electrons  are  nu- 
Insulotors  merous  and  loosely  held,  the 

element  is  a good  conductor. 
On  the  other  hand,  if  there  are  few  free  elec- 
trons, as  is  the  case  when  the  electrons  in  an 
outer  ring  are  tightly  held,  the  element  is  a 
poor  conductor.  If  there  are  virtually  no  free 
electrons,  the  element  is  a good  insulator. 

2-2  Fundamental  Electrical 
Units  and  Relatianships 

Electromotive  Force:  The  free  electrons  in 
Potential  Difference  a conductor  move  con- 
stantly about  and  change 
their  position  in  a haphazard  manner.  To 
produce  a drift  of  electrons  or  electric  current 
along  a wire  it  is  necessary  that  there  be  a 
difference  in  "pressure”  or  potential  between 
the  two  ends  of  the  wire.  This  potential  dif- 
ference can  be  produced  by  connecting  a 
source  of  electrical  potential  to  the  ends  of 
the  wire. 

As  will  be  explained  later,  there  is  an  ex- 
cess of  electrons  at  the  negative  terminal  of 
a battery  and  a deficiency  of  electrons  at  the 
positive  terminal,  due  to  chemical  action. 
When  the  battery  is  connected  to  the  wire,  the 
deficient  atoms  at  the  positive  terminal  attract 
free  electrons  from  the  wire  in  order  for  the 
positive  terminal  to  become  neutral.  The 
attracting  of  electrons  continues  through  the 
wire,  and  finally  the  excess  electrons  at 
the  negative  terminal  of  the  battery  are  at- 
tracted by  the  positively  charged  atoms  at  the 
end  of  the  wire.  Other  sources  of  electrical 
potential  (in  addition  to  a battery)  are:  an 
electrical  generator  (dynamo),  a thermocouple, 
an  electrostatic  generator  (static  machine),  a 
photoelectric  cell,  and  a crystal  or  piezo- 
electric generator. 

Thus  it  is  seen  that  a potential  difference 
is  the  result  of  a difference  in  the  number  of 
electrons  between  the  two  (or  more)  points  in 
question.  The  force  or  pressure  due  to  a 
potential  difference  is  termed  the  electro- 
motive force,  usually  abbreviated  e.m.f.  or 
E.  M.  F.  It  is  expressed  in  units  called  volts. 

It  should  be  noted  that  for  there  to  be  a 
potential  difference  between  two  bodies  or 
points  it  is  not  necessary  that  one  have  a 
positive  charge  and  the  other  a negative 
charge.  If  two  bodies  each  have  a negative 


charge,  but  one  more  negative  than  the  other, 
the  one  with  the  lesser  negative  charge  will 
act  as  though  it  were  positively  charged  with 
respect  to  the  other  body.  It  is  the  algebraic 
potential  difference  that  determines  the  force 
with  which  electrons  are  attracted  or  repulsed, 
the  potential  of  the  earth  being  taken  as  the 
zero  reference  point. 

The  Electric  The  flow  of  electrons  along  a 
Current  conductor  due  to  the  application 

of  an  electromotive  force  con- 
stitutes an  electric  current.  This  drift  is  in 
addition  to  the  irregular  movements  of  the 
electrons.  However,  it  must  not  be  thought 
that  each  free  electron  travels  from  one  end 
of  the  circuit  to  the  other.  On  the  contrary, 
each  free  electron  travels  only  a short  distance 
before  colliding  with  an  atom;  this  collision 
generally  knocking  off  one  or  more  electrons 
from  the  atom,  which  in  turn  move  a short 
distance  and  collide  with  other  atoms,  knock- 
ing off  other  electrons.  Thus,  in  the  general 
drift  of  electrons  along  a wire  carrying  an 
electric  current,  each  electron  travels  only  a 
short  distance  and  the  excess  of  electrons  at 
one  end  and  the  deficiency  at  the  other  are 
balanced  by  the  source  of  the  e.m.f.  When  this 
source  is  removed  the  state  of  normalcy  re- 
turns; there  is  still  the  rapid  interchange  of 
free  electrons  between  atoms,  but  there  is  no 
general  trend  or  "net  movement”  in  either 
one  direction  or  the  other. 

Ampere  and  There  are  two  units  of  measure- 
Coulomb  ment  associated  with  current, 

and  they  are  often  confused. 
The  rate  of  flow  of  electricity  is  stated  in 
amperes.  The  unit  of  quantity  is  the  coulomb. 
A coulomb  is  equal  to  6.28  x 10“  electrons, 
and  when  this  quantity  of  electrons  flows  by 
a given  point  in  every  second,  a current  of 
one  ampere  is  said  to  be  flowing.  An  ampere 
is  equal  to  one  coulomb  per  second;  a coulomb 
is,  conversely,  equal  to  one  ampere-second. 
Thus  we  see  that  coulomb  indicates  amount, 
and  ampere  indicates  rate  of  flow  of  electric 
current. 

Older  textbooks  speak  of  current  flow  as 
being  from  the  positive  terminal  of  the  e.m.f. 
source  through  the  conductor  to  the  negative 
terminal.  Nevertheless,  it  has  long  been  an 
established  fact  that  the  current  flow  in  a 
metallic  conductor  is  the  electronic  flow  from 
the  negative  terminal  of  the  source  of  voltage 
through  the  conductor  to  the  positive  terminal. 
The  only  exceptions  to  the  electronic  direction 
of  flow  occur  in  gaseous  and  electrolytic  con- 
ductors where  the  flow  of  positive  ions  toward 
the  cathode  or  negative  electrode  constitutes 
a positive  flow  in  the  opposite  direction  to  the 
electronic  flow.  (An  ion  is  an  atom,  molecule. 


HANDBOOK 


Resistance  23 


or  particle  which  either  lacks  one  or  more 
electrons,  or  else  has  an  excess  of  one  or 
more  electrons.) 

In  radio  work  the  terms  "electron  flow"  and 
"current”  are  becoming  accepted  as  being 
synonymous,  but  the  older  terminology  is  still 
accepted  in  the  electrical  (industrial)  field. 
Because  of  the  confusion  this  sometimes 
causes,  it  is  often  safer  to  refer  to  the  direc- 
tion of  electron  flow  rather  than  to  the  direc- 
tion of  the  "current.”  Since  electron  flow 
consisrs  actually  of  a passage  of  negative 
charges,  current  flow  and  algebraic  electron 
flow  do  pass  in  the  same  direction. 

Resistance  The  flow  of  current  in  a material 
depends  upon  the  ease  with 
which  electrons  can  be  detached  from  the 
atoms  of  the  material  and  upon  its  molecular 
structure.  In  other  words,  the  easier  it  is  to 
detach  electrons  from  the  atoms  the  more  free 
electrons  there  will  be  to  contribute  to  the 
flow  of  current,  and  the  fewer  collisions  that 
occur  between  free  electrons  and  atoms  the 
greater  will  be  the  total  electron  flow. 

The  opposition  to  a steady  electron  flow 
is  called  the  resistance  of  a material,  and  is 
one  of  its  physical  properties. 

The  unit  of  resistance  is  the  ohm.  Every 
substance  has  a specific  resistance,  usually 
expressed  as  ohms  per  mil-foot,  which  is  deter- 
mined by  the  material*  s molecular  structure 
and  temperature.  A mil-foot  is  a piece  of 
material  one  circular  mil  in  area  and  one  foot 
long.  Another  measure  of  resistivity  frequently 
used  is  expressed  in  the  units  microhms  per 
centimeter  cube.  The  resistance  of  a uniform 
length  of  a given  substance  is  directly  pro- 
portional to  its  length  and  specific  resistance, 
and  inversely  proportional  to  its  cross-section- 
al area.  A wire  with  a certain  resistance  for  a 
given  length  will  have  twice  as  much  resist- 
ance if  the  length  of  the  wire  is  doubled.  For 
a given  length,  doubling  the  cross-sectional 
area  of  the  wire  will  halve  the  resistance, 
while  doubling  the  diameter  will  reduce  the 
resistance  to  one  fourth.  This  is  true  since 
the  cross-sectional  area  of  a wire  varies  as 
the  square  of  the  diameter.  The  relationship 
between  the  resistance  and  the  linear  dimen- 
sions of  a conductor  may  be  expressed  by  the 
following  equation: 

r / 

R = — 

A 

Where 

R = resistance  in  ohms 
r=  resistivity  in  Ohms  per  mil- foot 
I = length  of  conductor  in  feet 

A = cross-sectional  area  in  circular  mils 


TABLE  OF  RESISTIVITY 

Moteriol 

Resistivity  in 
Ohms  per 
Circular 
Mil-Foot 

Temp.  CoeFf.  of 
resistance  per  °C 
at  20 » C. 

Aluminum 

17 

0.0049 

Brass 

45 

0.003  to  0.007 

Codmium 

46 

0.0038 

Chromium 

16 

0.00 

Copper 

10.4 

0.0039 

Iron 

59 

0.006 

Silver 

9.8 

0.004 

Zinc 

36 

0.0035 

Nichrome 

650 

0.0002 

Constonton 

295 

0.00001 

Manganin 

290 

0.00001 

Monel 

255 

0.0019 

FIGURE  1 


The  resistance  also  depends  upon  tempera- 
ture, increasing  with  increases  in  temperature 
for  most  substances  (including  most  metals), 
due  to  increased  electron  acceleration  and 
hence  a greater  number  of  impacts  between 
electrons  and  atoms.  However,  in  the  case  of 
some  substances  such  as  carbon  and  glass  the 
temperature  coefficient  is  negative  and  the 
resistance  decreases  as  the  temperature  in- 
creases. This  is  also  true  of  electrolytes.  The 
temperature  may  be  raised  by  the  external  ap- 
plication of  heat,  or  by  the  flow  of  the  current 
itself.  In  the  latter  case,  the  temperature  is 
raised  by  the  heat  generated  when  the  electrons 
and  atoms  collide. 

Conductors  and  In  the  molecular  structure  of 
Insulators  many  materials  such  as  glass, 

porcelain,  and  mica  all  elec- 
trons are  tightly  held  within  their  orbits  and 
there  are  comparatively  few  free  electrons. 
This  type  of  substance  will  conduct  an  elec- 
tric current  only  with  great  difficulty  and  is 
known  as  an  insulator.  An  insulator  is  said  to 
have  a high  electrical  resistance. 

On  the  other  hand,  materials  that  have  a 
large  number  of  free  electrons  are  known  as 
conductors.  Most  metals,  those  elements  which 
have  only  one  or  two  electrons  in  their  outer 
ring,  are  good  conductors.  Silver,  copper,  and 
aluminum,  in  that  order,  are  the  best  of  the 
common  metals  used  as  conductors  and  are 
said  to  have  the  greatest  conductivity,  or  low- 
est resistance  to  the  flow  of  an  electric 
current. 

Fundamental  These  units  ate  the  volt, 

Electrical  Units  the  ampere,  and  the  ohm. 

They  were  mentioned  in  the 
preceding  paragraphs,  but  were  not  completely 
defined  in  terms  of  fixed,  known  quantities. 

The  fundamental  unit  of  current,  or  rate  of 
flow  of  electricity  is  the  ampere.  A current  of 
one  ampere  will  deposit  silver  from  a speci- 
fied solution  of  silver  nitrate  at  a rate  of 
1.118  milligrams  per  second. 


24  Direct  Current  Circuits 


THE  RADIO 


Figure  2 

TYPICAL  RESISTORS 

Shown  above  are  various  types  of  resistor*  used  In  electronic  circuits.  The  larger  units  are 
power  resistors.  On  the  left  Is  a variable  power  resistor.  Three  preelslon^type  resistors  are 
shown  in  the  center  with  two  small  composit/on  resistors  beneath  them.  At  the  right  Is  a 
composition>type  potentiometer,  used  for  audio  circuitry. 


The  international  standard  for  the  ohm  is 
the  resistance  offered  by  a uniform  column  of 
mercury  at  0°C.,  14.4521  grams  in  mass,  of 
constant  cross-sectional  area  and  106.300 
centimeters  in  length.  The  expression  megohm 
(1,000,000  ohms)  is  also  sometimes  used 
when  speaking  of  very  large  values  of  resist- 
ance. 

A volt  is  the  e.m.f.  that  will  produce  a cur- 
rent of  one  ampere  through  a resistance  of 
one  ohm.  The  standard  of  electromotive  force 
is  the  Weston  cell  which  at  20°  C.  has  a 
potential  of  1.0183  volts  across  its  terminals. 
This  cell  is  used  only  for  reference  purposes 
in  a bridge  circuit,  since  only  an  infinitesimal 


amount  of  current  may  be  drawn  from  it  with- 
out disturbing  its  characteristics. 

Ohm’s  Law  The  relationship  between  the 
electromotive  force  (voltage), 
the  flow  of  current  (amperes),  and  the  resist- 
ance which  impedes  the  flow  of  current  (ohms), 
is  very  clearly  expressed  in  a simple  but 
highly  valuable  law  known  as  Ohm’s  law. 
This  law  states  that  the  current  in  amperes  is 
equal  to  the  voltage  in  volts  divided  by  the 
resistance  in  ohms.  Expressed  as  an  equation: 

E 

I = - 

R 


RESISTANCE 


I ■- — CONDUCTORS  — 


BATTERY 


t 


Rl  — ► Rz 

- -VW—  — 


E 


Figure  3 

SIMPLE  SERIES  CIRCUITS 

At  (A)  the  battery  is  in  series  with  a single 
resistor.  At  (B)  the  battery  is  in  series  with 
two  resistors,  the  resistors  themselves  being 
in  series.  The  arrows  indicote  the  direction  of 
electron  flow. 


If  the  voltage  (E)  and  resistance  (R)  are 
known,  the  current  (I)  can  be  readily  found. 
If  the  voltage  and  current  are  known,  and  the 
resistance  is  unknown,  the  resistance  (R)  is 

E 

equal  to  — . When  the  voltage  is  the  un- 
known quantity,  it  can  be  found  by  multiply- 
ing I X R.  These  three  equations  are  all  secured 
from  the  original  by  simple  transposition. 
The  egressions  are  here  repeated  for  quick 
reference; 

E E 

I=—  R-—  E=IR 

R I 


HANDBOOK 


Resistive  Circuits  25 


Figure  4 

SIMPLE  PARALLEL 
CIRCUIT 


Figure  5 

SERIES-PARALLEL 

CIRCUIT 


The  two  res/sfors  Rj  anti  R2  are  sold  to  bo  in 
parallol  since  tho  flow  of  currant  is  offorad 
fwo  parallol  paths.  An  oloctran  loaving  point 
A will  pass  oltbor  through  or  R2,  but  not 
through  both,  to  roach  tho  positivo  tormina! 
of  tho  battory.  If  a largo  numbor  of  oloctrons 
are  considered,  tho  groator  numbor  will  pass 
thraugh  wblehovor  of  tho  two  rosistors  has 
tho  lower  rosistanco. 


where  / is  the  current  in  amperes, 

K is  the  resistance  in  ohms, 

E is  the  electromotive  force  in  volts. 


Application  of  All  electrical  circuits  fall  in- 
Ohm's  Low  to  one  of  three  classes;  series 

circuits,  parallel  circuits,  and 
series-parallel  circuits.  A series  circuit  is 
one  in  which  the  current  flows  in  a single 
continuous  path  and  is  of  the  same  value  at 
every  point  in  the  circuit  (figure  3).  In  a par- 
allel circuit  there  are  two  or  more  current 
paths  between  two  points  in  the  circuit,  as 
shown  in  figure  4.  Here  the  current  divides  at 
A,  part  going  through  R,  and  part  through  Rj, 
and  combines  at  B to  return  to  the  battery. 
Figure  5 shows  a series-parallel  circuit.  There 
are  two  paths  between  points  A and  B as  in 
the  parallel  circuit,  and  in  addition  there  are 
two  resistances  in  series  in  each  branch  of 
the  parallel  combination.  Two  other  examples 
of  series-parallel  arrangements  appear  in  fig- 
ure 6.  The  way  in  which  the  current  splits  to 
flow  through  the  parallel  branches  is  shown  by 
the  arrows. 

In  every  circuit,  each  of  the  parts  has  some 
resistance;  the  batteries  or  generator,  the  con- 
necting conductors,  and  the  apparatus  itself. 
Thus,  if  each  part  has  some  resistance,  no 
matter  how  little,  and  a current  is  flowing 
through  it,  there  will  be  a voltage  drop  across 
it.  In  other  words,  there  will  be  a potential 
difference  between  the  two  ends  of  the  circuit 
element  in  question.  This  drop  in  voltage  is 
equal  to  the  product  of  the  current  and  the 
resistance,  hence  it  is  called  the  IR  drop. 

The  source  of  voltage  has  an  internal  re- 
sistance, and  when  connected  into  a circuit 
so  that  current  flows,  there  will  be  an  IR  drop 
in  the  source  just  as  in  every  other  part  of  the 
circuit.  Thus,  if  the  terminal  voltage  of  the 
source  could  be  measured  in  a way  that  would 
cause  no  current  to  flow,  it  would  be  found 
to  be  mote  than  the  voltage  measured  when  a 
current  flows  by  the  amount  of  the  IR  drop 


In  this  type  of  circuit  the  resistors  are  ar- 
ranged In  series  groups,  and  these  serlesed 
groups  are  then  pieced  in  parallel. 


in  the  source.  The  voltage  measured  with  no 
current  flowing  is  termed  the  no  load  voltage; 
that  measured  with  current  flowing  is  the  load 
voltage.  It  is  apparent  that  a voltage  source 
having  a low  internal  resistance  is  most  de- 
sirable. 


Resistances  The  current  flowing  in  a series 
in  Series  circuit  is  equal  to  the  voltage 

impressed  divided  by  the  total 
resistance  across  which  the  voltage  is  im- 
pressed. Since  the  same  current  flows  through 
every  part  of  the  circuit,  it  is  merely  nec- 
essary to  add  all  the  individual  resistances  to 
obtain  the  total  resistance.  Expressed  as  a 
formula; 

^^total  = Ri  + R2  + Rj  + • • • + Rn  • 

Of  course,  if  the  resistances  happened  to  be 
all  the  same  value,  the  total  resistance  would 
be  the  resistance  of  one  multiplied  by  the 
number  of  resistors  in  the  circuit. 


Resi;stances  Consider  two  resistors,  one  of 
in  Parallel  100  ohms  and  one  of  10  ohms, 
connected  in  parallel  as  in  fig- 
ure 4,  with  a voltage  of  10  volts  applied 
across  each  resistor,  so  the  current  through 
each  can  be  easily  calculated. 

E 

I = — 

R 


E = 10  volts 
R = 100  ohms 

E = 10  volts 
R = 10  ohms 


10 

=0.1  ampere 

100 

10 

— =1.0  ampere 
10 


Total  current  = 1,  -H  Ij  = 1.1  ampere 

Until  it  divides  at  A,  the  entire  current  of 
1.1  amperes  is  flowing  through  the  conductor 
from  the  battery  to  A,  and  again  from  B through 
the  conductot  to  the  battery.  Since  this  is  more 
current  than  flows  through  the  smaller  resistor 
it  is  evident  that  the  resistance  of  the  parallel 
combination  must  be  less  than  10  ohms,  the 
resistance  of  the  smaller  resistor.  We  can  find 
this  value  by  applying  Ohm’s  law. 


26  Direct  Current  Circuits 


THE  RADIO 


E = 10  volts 
1=1.1  amperes 


10 

R = = 9-09  ohms 

1.1 


The  resistance  of  the  parallel  combination  is 
9.09  ohms. 

Mathematically,  we  can  derive  a simple 
formula  for  finding  the  effective  resistance  of 
two  resistors  connected  in  parallel. 

This  formula  is: 


A 


@ 


A 


(D 


XV  = 

R, +R, 

where  R is  the  unknown  resistance, 

Ri  is  the  resistance  of  the  first  resistor, 
R,  is  the  resistance  of  the  second  re- 
sistor. 

If  the  effective  value  required  is  known, 
and  it  is  desired  to  connect  one  unknown  re- 
sistor in  parallel  with  one  of  known  value, 
the  following  transposition  of  the  above  for- 
mula will  simplify  the  problem  of  obtaining 
the  unknown  value: 

Ri  X R 


where  R is  the  effective  value  required, 

Rj  is  the  known  resistor, 

Rj  is  the  value  of  the  unknown  resist- 
ance necessary  to  give  R when 
in  parallel  with  Rj. 

The  resultant  value  of  placing  a number  of 
unlike  resistors  in  parallel  is  equal  to  the  re- 
ciprocal of  the  sum  of  the  reciprocals  of  the 
various  resistors.  This  can  be  expressed  as: 

1 

R 

111  1 

— + — + — + . ...  — 

R,  Rj  Rj  Rn 

The  effective  value  of  placing  any  number 
of  unlike  resistors  in  parallel  can  be  deter- 
mined from  the  above  formula.  However,  it 
is  commonly  used  only  when  there  are  three 
or  more  resistors  under  consideration,  since 
the  simplified  formula  given  before  is  more 
convenient  when  only  two  resistors  are  being 
used. 

From  the  above,  it  also  follows  that  when 
two  or  more  resistors  of  the  same  value  are 
placed  in  parallel,  the  effective  resistance  of 
the  paralleled  resistors  is  equal  to  the  value 
of  one  of  the  resistors  divided  by  the  number 
of  resistors  in  parallel. 

The  effective  value  of  resistance  of  two  or 


Figure  6 

OTHER  COMMON  SERIES-PARALLEL 
CIRCUITS 


more  resistors  connected  in  parallel  is  always 
less  than  the  value  of  the  lowest  resistance  in 
the  combination.  It  is  well  to  bear  this  simple 
rule  in  mind,  as  it  will  assist  greatly  in  ap- 
proximating the  value  of  paralleled  resistors. 

Resistors  in  To  find  the  total  resistance  of 
Series  Parallel  several  resistors  connected  in 
series-parallel,  it  is  usually 
easiest  to  apply  either  the  formula  for  series 
resistors  or  the  parallel  resistor  formula  first, 
in  order  to  reduce  the  original  arrangement  to 
a simpler  one.  For  instance,  in  figure  5 the 
series  resistors  should  be  added  in  each 
branch,  then  there  will  be  but  two  resistors  in 
parallel  to  be  calculated.  Similarly  in  figure  7, 
although  here  there  will  be  three  parallel  re- 
sistors after  adding  the  series  resistors  in 
each  branch.  In  figure  6B  the  paralleled  re- 
sistors should  be  reduced  to  the  equivalent 
series  value,  and  then  the  series  resistance 
values  can  be  added. 

Resistances  in  series-parallel  can  be  solved 
by  combining  the  series  and  parallel  formulas 
into  one  similar  to  the  following  (refer  to 
figure  7): 

1 

R = 

1 1 1 

+ h 

Ri  + Rj  R3  + R4  R5  + Rg  Ry 

Valtoge  Dividers  A voltage  divider  is  ex- 
actly what  its  name  im- 
plies: a resistor  or  a series  of  resistors  con- 
nected across  a source  of  voltage  from  which 
various  lesser  values  of  voltage  may  be  ob- 
tained by  connection  to  various  points  along 
the  resistor. 

A voltage  divider  serves  a most  useful  pur- 
pose in  a radio  receiver,  transmitter  or  ampli- 
fier, because  it  offers  a simple  means  of 
obtaining  plate,  screen,  and  bias  voltages  of 
different  values  from  a common  power  supply 


HANDBOOK 


Voltage  Divider  27 


Sr  1 

Sr3 

Irs 

T 

f Re 

R 

k^2 

|R4 

Ir? 

_L 

Figure  7 

ANOTHER  TYPE  OF 
SERIES-PARALLEL  CIRCUIT 


source.  It  may  also  be  used  to  obtain  very  low 
voltages  of  the  order  of  .01  to  .001  volt  with 
a high  degree  of  accuracy,  even  though  a 
means  of  measuring  such  voltages  is  lacking. 
The  procedure  for  making  these  measurements 
can  best  be  given  in  the  following  example. 

Assume  that  an  accurately  calibrated  volt- 
meter reading  from  0 to  150  volts  is  available, 
and  that  the  source  of  voltage  is  exactly  100 
volts.  This  100  volts  is  then  impressed  through 
a resistance  of  exactly  1,000  ohms.  It  will, 
then,  be  found  that  the  voltage  along  various 
points  on  the  resistor,  with  respect  to  the 
grounded  end,  is  exactly  proportional  to  the 
resistance  at  that  point.  From  Ohm’s  law,  the 
current  would  be  0.1  ampere;  this  current  re- 
mains unchanged  since  the  original  value  of 
resistance  (1,000  ohms)  and  the  voltage  source 
(100  volts)  ate  unchanged.  Thus,  at  a 500- 
ohm  point  on  the  resistor  (half  its  entire  re- 
sistance), the  voltage  will  likewise  be  halved 
or  reduced  to  50  volts. 

The  equation  (E  = I x R)  gives  the  proof; 
E = 500  X 0.1  = 50.  At  the  point  of  250  ohms 
on  the  resistor,  the  voltage  will  be  one-fourth 
the  total  value,  or  25  volts  (E  = 250  x 0.1  = 25). 
Continuing  with  this  process,  a point  can  be 
found  where  the  resistance  measures  exactly 
1 ohm  and  where  the  voltage  equals  0.1  volt. 
It  is,  therefore,  obvious  that  if  the  original 
source  of  voltage  and  the  resistance  can  be 
measured,  it  is  a simple  matter  to  predeter- 
mine the  voltage  at  any  point  along  the  resist- 
or, provided  that  the  current  remains  constant, 
and  provided  that  no  current  is  taken  from  the 
tap-on  point  unless  this  current  is  taken  into 
consideration. 

Voltage  Divider  Proper  design  of  a voltage 
Calculations  divider  for  any  type  of  radio 
equipment  is  a relatively 
simple  matter.  The  first  consideration  is  the 
amount  of  "bleeder  current”  to  be  drawn. 
In  addition,  it  is  also  necessary  that  the  de- 
sired voltage  and  the  exact  current  at  each  tap 
on  the  voltage  divider  be  known. 

Figure  8 illustrates  the  flow  of  current  in  a 
simple  voltage  divider  and  load  circuit.  The 
light  arrows  indicate  the  flow  of  bleeder  cur- 
rent, while  the  heavy  arrows  indicate  the  flow 
of  the  load  current.  The  design  of  a combined 


iA 


BLEEDER  CURRENT  ; 
FLOWS  BETWEEN 
POINTS  A AND  B ■ 

>•  1 

• ♦ EXTERNAL 

I LOAD 

B 

Figure  8 

SIMPLE  VOLTAGE  DIVIDER 
CIRCUIT 

The  arreWM  Indicate  the  manner  In  which  the 
current  flow  dtvtdea  between  the  voltage  divider 
Itself  and  the  external  load  circuit. 


bleeder  resistor  and  voltage  divider,  such  as 
is  commonly  used  in  radio  equipment,  is  illus- 
trated in  the  following  example: 

A power  supply  delivers  300  volts  and  is 
conservatively  rated  to  supply  all  needed  cur- 
rent for  the  receiver  and  still  allow  a bleeder 
current  of  10  milliamperes.  The  following  volt- 
ages ate  wanted:  75  volts  at  2 milliamperes 
for  the  detector  tube,  100  volts  at  5 milli- 
amperes for  the  screens  of  the  tubes,  and 
250  volts  at  20  milliamperes  for  the  plates  of 
the  tubes.  The  required  voltage  drop  across  R, 
is  75  volts,  across  Rj  25  volts,  across  Rj  150 
volts,  and  across  R,  it  is  50  volts.  These 
values  are  shown  in  the  diagram  of  figure  9. 
The  respective  current  values  ate  also  indi- 
cated. Apply  Ohm’s  law: 

E 75 

R,  = — = = 7,500  ohms. 

I .01 

E 25 

R,  = — = = 2,083  ohms. 

I .012 

E 150 

R,  = — = = 8,823  ohms. 

I .017 

E 50 

R,  = — = = 1,351  ohms. 

I .037 

^Totei  = 7,500  -I-  2,083  -i-  8,823  -i 

1,351  = 19,757  ohms. 

A 20,000-ohm  resistor  with  three  sliding  taps 
will  be  of  the  approximately  correct  size,  and 
would  ordinarily  be  used  because  of  the  diffi- 
culty in  securing  four  separate  resistors  of  the 
exact  odd  values  indicated,  and  because  no 
adjustment  would  be  possible  to  compensate 
for  any  slight  error  in  estimating  the  probable 
currents  through  the  various  taps. 

When  the  sliders  on  the  resistor  once  are 
set  to  the  proper  point,  as  in  the  above  ex- 


28  Direct  Current  Circuits 


THE  RADIO 


Figure  9 

MORE  COMPLEX  VOLTAGE  DIVIDER 

The  method  for  computing  the  values  of  the 
resistors  Is  discussed  In  the  accompanying  test. 


ample,  the  voltages  will  remain  constant  at 
the  values  shown  as  long  as  the  current  re- 
mains a constant  value. 

Disadvantages  of  One  of  the  serious  disadvan- 
Voltage  Dividers  tages  of  the  voltage  divider 
becomes  evident  when  the 
the  current  drawn  fromone  of  the  taps  changes. 
It  is  obvious  that  the  voltage  drops  are  inter- 
dependent and,  in  turn,  the  individual  drops 
are  in  proportion  to  the  current  which  flows 
through  the  respective  sections  of  the  divider 
resistor.  The  only  remedy  lies  in  providing  a 
heavy  steady  bleeder  current  in  order  to  make 
the  individual  currents  so  small  a part  of  the 
total  current  that  any  change  in  current  will 
result  in  only  a slight- change  in  voltage.  This 
can  seldom  be  realized  in  practice  because  of 
the  excessive  values  of  bleeder  current  which 
would  be  required. 

Kirchhoff’s  Laws  Ohm’s  law  is  all  that  is 
necessary  to  calculate  the 
values  in  simple  circuits,  such  as  the  pre- 
ceding examples;  but  in  mote  complex  prob- 
lems, involving  several  loops  or  more  than 
one  voltage  in  the  same  closed  circuit,  the 
use  of  Kirchhojj’ s laws  will  greatly  simplify 
the  calculations.  These  laws  are  merely  rules 
for  applying  Ohm’s  law. 

Kirchhoff’s  first  law  is  concerned  with  net 
current  to  a point  in  a circuit  and  states  that: 

At  any  point  in  a circuit  the  current 
flowing  toward  the  point  is  equal  to 
the  current  flowing  away  from  the 
point. 

Stated  in  another  way:  if  currents  flowing  to 
the  point  ate  considered  positive,  and  those 
flowing  from  the  point  are  considered  nega- 


—  9 AUPC 


20  volts 


Figure  10 

ILLUSTRATING  KIRCHHOFF’S 
FIRST  LAW 

The  current  flowing  toward  point  "A”  Is  equal 
to  the  current  flowing  away  from  point  "A.” 


tive,  the  sum  of  all  currents  flowing  toward 
and  away  from  the  point  — taking  signs  into 
account  — is  equal  to  zero.  Such  a sum  is 
known  as  an  algebraic  sum;  such  that  the  law 
can  be  stated  thus:  The  algebraic  sum  of  all 
currents  entering  and  leaving  a point  is  zero. 

Figure  10  illustrates  this  first  law.  Since  the 
effective  tesistance  of  the  network  of  resistors 
is  5 ohms,  it  can  be  seen  that  4 amperes  flow 
toward  point  A,  and  2 amperes  flow  away 
through  the  two  5-ohm  resistors  in  series.  The 
remaining  2 amperes  flow  away  through  the  10- 
ohm  resistor.  Thus,  there  are  4 amperes  flowing 
to  point  A and  4 amperes  flowing  away  from 
the  point.  If  R is  the  effective  resistance  of 
the  network  (5  ohms),  R,  = 10  ohms,  Rj  = 5 
ohms,  R3  = 5 ohms,  and  E = 20  volts,  we  can 
set  up  the  following  equation: 

EE  E 

R Rj  R2  'f'  R3 

20  20  20 

5 10  5 -I-  5 

4 -2  —2  = 0 

Kirchhoff’s  second  law  is  concerned  with 
net  voltage  drop  around  a closed  loop  in  a 
circuit  and  states  that: 

In  any  closed  path  or  loop  in  a circuit 

the  sum  of  the  IR  drops  must  equal 
the  sum  of  the  applied  e.m.f.’s. 

The  second  law  also  may  be  conveniently 
stated  in  terms  of  an  algebraic  sum  as:  The 
algebraic  sum  of  all  voltage  drops  around  a 
closed  path  or  loop  in  a circuit  is  zero.  The 
applied  e.m.f.’s  (voltages)  are  considered 
positive,  while  IR  drops  taken  in  the  direction 
of  current  flow  (including  the  internal  drop 
of  the  sources  of  voltage)  are  considered 
negative. 

Figure  11  shows  an  example  of  the  applica- 
tion of  Kirchhoff’s  laws  to  a comparatively 
simple  circuit  consisting  of  three  resistors  and 


HANDBOOK 


Kirchoff's  Laws  29 


1.  SET  VOLTAGE  DROPS  AROUND  EACH  LOOP  EQUAL  TO  ZERO. 
Ii  2(ohms)'^2  (l-i  -I2)  +3=  0 (first  loop) 

-6  + 2 (l2  - r 1)  +3  l2  = 0 (second  loop) 


2,  simplify 

211  + 211-212+3  = 0 2U- 211+312-6=0 

4I1+3  _ , 512-211-6  =0 

2 ' 2Ii  +6  , 

—5 — 'I2 

3,  EQUATE 

411  + 3 ^ 211  + 6 
2 5 


Figure  1 1 

ILLUSTRATING  KIRCHHOFF’S 
SECOND  LAW 

T/ie  vo/fege  drop  around  any  elotod  loop  In  a 
notwork  Is  equal  to  zero. 


two  batteries.  First  assume  an  arbitrary  direc- 
tion of  current  flow  in  each  closed  loop  of  the 
circuit,  drawing  an  arrow  to  indicate  the  as- 
sumed direction  of  current  flow.  Then  equate 
the  sum  of  all  IR  drops  plus  battery  drops 
around  each  loop  to  zero.  You  will  need  one 
equation  for  each  unknown  to  be  determined. 
Then  solve  the  equations  for  the  unknown  cur- 
rents in  the  general  manner  indicated  in  figure 
11.  If  the  answer  comes  out  positive  the  di- 
rection of  current  flow  you  originally  assumed 
was  correct.  If  the  answer  comes  out  negative, 
the  current  flow  is  in  the  opposite  direction  to 
the  arrow  which  was  drawn  originally.  This  is 
illustrated  in  the  example  of  figure  II  where 
the  direction  of  flow  of  Ij  is  opposite  to  the 
direction  assumed  in  the  sketch. 

Power  in  In  order  to  cause  electrons 

Resistive  Circuits  to  flow  through  a conductor, 
constituting  a current  flow, 
it  is  necessary  to  apply  an  electromotive  force 
(voltage)  across  the  circuit.  Less  power  is 
expended  in  creating  a small  current  flow 
through  a given  resistance  than  in  creating 
a large  one;  so  it  is  necessary  to  have  a unit 
of  power  as  a reference. 

The  unit  of  electrical  power  is  the  watt, 
which  is  the  rate  of  energy  consumption  when 


an  e.m.f.  of  1 volt  forces  a current  of  1 ampere 
through  a circuit.  The  power  in  a resistive 
circuit  is  equal  to  the  product  of  the  volt- 
age applied  across,  and  the  current  flowing 
in,  a given  circuit.  Hence:  P (watts)  = E 
(volts)  X I (amperes). 

Since  it  is  often  convenient  to  express 
power  in  terms  of  the  resistance  of  the  circuit 
and  the  current  flowing  through  it,  a substi- 
tution of  IR  for  E (E  A=  IR)  in  the  above  formula 
gives:  P = IR  X I or  P = I^R.  In  terms  of  volt- 
age and  resistance,  P = EVR.  Here,  I = E/R 
and  when  this  is  substituted  for  I the  original 
formula  becomes  P = E X E/R,  or  P = EVR. 
To  repeat  these  three  expressions: 

P = El,  P = I^R,  and  P = EVR, 

where  P is  the  power  in  watts, 

E is  the  electromotive  force  in  volts, 
and 

I is  the  current  in  amperes. 

To  apply  the  above  equations  to  a typical 
problem:  The  voltage  drop  across  a cathode 
resistor  in  a power  amplifier  stage  is  50  volts; 
the  plate  current  flowing  through  the  resistor 
is  150  milliamperes.  The  number  of  watts  the 
resistor  will  be  required  to  dissipate  is  found 
from  the  formula:  P = El,  or  50  x .150  = 7.5 
watts  (.150  amperes  is  equal  to  150  milli- 
amperes). From  the  foregoing  it  is  seen  that 
a 7.5-watt  resistor  will  safely  carry  the  re- 
quired current,  yet  a 10-  or  20-watt  resistor 
would  ordinarily  be  used  to  provide  a safety 
factor. 

In  another  problem,  the  conditions  being 
similar  to  those  above,  but  with  the  resistance 
(R  = 333*/  ohms),  and  current  being  the  known 
factors,  the  solution  is  obtained  as  follows: 
P = I^R  = .0225  X 333.33  = 7.5.  If  only  the  volt- 
age and  resistance  are  known,  P = EVR  = 
2500/333.33  = 7.5  watts.  It  is  seen  that  all 
three  equations  give  the  same  results;  the 
selection  of  the  particular  equation  depends 
only  upon  the  known  factors. 

Power,  Energy  It  is  important  to  remember 
and  Work  that  power  (expressed  in  watts, 

horsepower,  etc.),  represents 
the  rate  of  energy  consumption  or  the  rate  of 
doing  work.  But  when  we  pay  our  electric  bill 


Figure  12 
MATCHING  OF 
RESISTANCES 

To  deliver  the  greatest  amount  of  power  to  the 
load,  the  load  resistance  should  be  equal  to 
the  Internal  resistance  of  the  battery  Rj . 


30  Direct  Current  Circuits 


THE  RADIO 


Figure  13 

TYPICAL  CAPACITORS 

T/ie  rwo  large  units  are  high  value  flltnr  capaci- 
tors. Shawn  hanaath  thasa  arm  various  typos  of 
hy-pass  capacitors  for  r-f  and  audio  applicaflon. 


to  the  power  company  we  have  purchased  a 
specific  amount  of  energy  or  work  expressed 
in  the  common  units  of  kilowatt-hours.  Thus 
rate  of  energy  consumption  (watts  or  kilowatts) 
multiplied  by  time  (seconds,  minutes  or  hours) 
gives  us  total  energy  or  work.  Other  units  of 
energy  are  the  watt-second,  BTU,  calorie,  erg, 
and  joule. 

Heating  Effect  Heat  is  generated  when  a 
source  of  voltage  causes  a 
current  to  flow  through  a resistor  (or,  for  that 
matter,  through  any  conductor).  As  explained 
earlier,  this  is  due  to  the  fact  that  heat  is 
given  off  when  free  electrons  collide  with  the 
atoms  of  the  material.  More  heat  is  generated 
in  high  resistance  materials  than  in  those  of 
low  resistance,  since  the  free  electrons  must 
strike  the  atoms  harder  to  knock  off  other 
electrons.  As  the  heating  effect  is  a function 
of  the  current  flowing  and  the  resistance  of 
the  circuit,  the  power  expended  in  heat  is 
given  by  the  second  formula;  P = I’R. 


2-3  Electrostatics  — Capacitors 

Electrical  energy  can  be  stored  in  an  elec- 
trostatic field.  A device  capable  of  storing 
energy  in  such  a field  is  called  capacitor 
(in  earlier  usage  the  term  condenser  was 
frequently  used  but  the  IRE  standards  call  for 
the  use  of  capacitor  instead  of  condenser)  and 


is  said  to  have  a certain  capacitance.  The 
energy  stored  in  an  electrostatic  field  is  ex- 
pressed in  joules  (watt  seconds)  and  is  equal 
to  CE’/2,  where  C is  the  capacitance  in  farads 
(a  unit  of  capacitance  to  be  discussed)  and  E 
is  the  potential  in  volts.  The  charge  is  equal 
to  CE,  the  charge  being  expressed  in  coulombs. 

Capacitance  and  Two  metallic  plates  sep- 
Capacitors  arated  from  each  other  by 

a thin  layer  of  insulating 
material  (called  a dielectric,  in  this  case), 
becomes  a capacitor.  When  a source  of  d-c 
potential  is  momentarily  applied  across  these 
plates,  they  may  be  said  to  become  charged. 
If  the  same  two  plates  are  then  joined  to- 
gether momentarily  by  means  of  a switch,  the 
capacitor  will  discharge. 

When  the  potential  was  first  applied,  elec- 
trons immediately  flowed  from  one  plate  to  the 
other  through  the  battery  or  such  source  of 
d-c  potential  as  was  applied  to  the  capacitor 
plates.  However,  the  circuit  from  plate  to 
plate  in  the  capacitor  was  incomplete  (the  two 
plates  being  separated  by  an  insulator)  and 
thus  the  electron  flow  ceased,  meanwhile  es- 
tablishing a shortage  of  electrons  on  one  plate 
and  a surplus  of  electrons  on  the  other. 

Remember  that  when  a deficiency  of  elec- 
trons exists  at  one  end  of  a conductor,  there 
is  always  a tendency  for  the  electrons  to  move 
about  in  such  a manner  as  to  re-establish  a 
state  of  balance.  In  the  case  of  the  capacitor 
herein  discussed,  the  surplus  quantity  of  elec- 
trons on  one  of  the  capacitor  plates  cannot 
move  to  the  other  plate  because  the  circuit 
has  been  broken;  that  is,  the  battery  or  d-c  po- 
tential was  removed.  This  leaves  the  capaci- 
tor in  a charged  condition;  the  capacitor  plate 
with  the  electron  deficiency  is  positively 
charged,  the  other  plate  being  negative. 

In  this  condition,  a considerable  stress 
exists  in  the  insulating  material  (dielectric) 
which  separates  the  two  capacitor  plates,  due 
to  the  mutual  attraction  of  two  unlike  poten- 
tials on  the  plates.  This  stress  is  known  as 
electrostatic  energy,  as  contrasted  with  elec- 
tromagnetic energy  in  the  case  of  an  inductor. 
This  charge  can  also  be  called  potential 
energy  because  it  is  capable  of  performing 
work  when  the  charge  is  released  through  an 
external  circuit.  The  charge  is  proportional  to 
the  voltage  but  the  energy  is  proportional  to 
the  voltage  squared,  as  shown  in  the  following 
analogy. 

The  charge  represents  a definite  amount  of 
electricity,  or  a given  number  of  electrons. 
The  potential  energy  possessed  by  these 
electrons  depends  not  only  upon  their  number, 
but  also  upon  their  potential  or  voltage. 

Compare  the  electrons  to  water,  and  two 
capacitors  to  standpipes,  a 1 pfd.  capacitor  to 


HANDBOOK 


Capacitance  31 


SURPLUS 
OF  ELECTRONS 


CHARGING  CURRENT 


Figure  14 

SIMPLE  CAPACITOR 

Illustrating  the  Imaginary  llnas  af  fores  rsprs- 
ssnting  ths  paths  along  which  ths  rspslling  fores 
of  ths  sisetrons  would  act  on  a frss  sisetron 
loeatsd  hstwssn  ths  two  capacitor  platss. 


a Standpipe  having  a cross  section  of  1 square 
inch  and  a 2 pfd.  capacitor  to  a standpipe  hav- 
ing a cross  section  of  2 square  inches.  The 
charge  will  represent  a given  volume  of  water, 
as  the  "charge”  simply  indicates  a certain 
number  of  electrons.  Suppose  the  water  is 
equal  to  5 gallons. 

Now  the  potential  energy,  or  capacity  for 
doing  work,  of  the  5 gallons  of  water  will  be 
twice  as  great  when  confined  to  the  1 sq.  in. 
standpipe  as  when  confined  to  the  2 sq.  in. 
standpipe.  Yet  the  volume  of  water,  or  "charge” 
is  the  same  in  either  case. 

Likewise  a 1 ^fd.  capacitor  charged  to  1000 
volts  possesses  twice  as  much  potential 
energy  as  does  a 2 pfd.  capacitor  charged  to 
500  volts,  though  the  charge  (expressed  in 
coulombs:  Q = CE)  is  the  same  in  either  case. 

The  Unit  of  Copoc-  If  the  external  circuit  of 
itance;  The  Farad  the  two  capacitor  plates  is 
completed  by  joining  the 
terminals  together  with  a piece  of  wire,  the 
electrons  will  rush  immediately  from  one  plate 
to  the  other  through  the  external  circuit  and 
establish  a state  of  equilibrium.  This  latter 
phenomenon  explains  the  discharge  of  a capac- 
itor. The  amount  of  stored  energy  in  a charged 
capacitor  is  dependent  upon  the  charging  po- 
tential, as  well  as  a factor  which  takes  into 
account  the  size  of  the  plates,  dielectric 
thickness,  nature  of  the  dielectric,  and  the 
number  of  plates.  This  factor,  which  is  de- 
termined by  the  foregoing,  is  called  the  capaci- 
tanceof  a capacitor  andis  expressed  in  farads. 

The  farad  is  such  a large  unit  of  capaci- 
tance that  it  is  rarely  used  in  radio  calcula- 
tions, and  the  following  more  practical  units 
have,  therefore,  been  chosen. 

1 «tcro/(*raii  = 1/1,000,000  of  a farad,  or 

.000001  farad,  or  10“®  farads. 


1 micro-microfarad  = 1/1,000,000  of  a micro- 
farad, or  .000001  microfarad,  or  10“®  mi- 
crofarads, 

1 micro- micro  farad  = one-millionth  of  one- 
millionth  of  a farad,  or  10“*’  farads. 


If  the  capacitance  is  to  be  expressed  in 
microfarads  in  the  equation  given  for  energy 
storage,  the  factor  C would  then  have  to  be 
divided  by  1,000,000,  thus; 

C X E’ 

Stored  energy  in  joules  = 

2 X 1,000,000 

This  storage  of  energy  in  a capacitor  is  one 
of  its  very  important  properties,  particularly 
in  those  capacitors  which  are  used  in  power 
supply  filter  circuits. 

Dielectric  Although  any  substance  which  has 
Materials  the  characteristics  of  a good  in- 
sulator may  be  used  as  a dielec- 
tric material,  commercially  manufactured  ca- 
pacitors make  use  of  dielectric  materials 
which  have  been  selected  because  their  char- 
acteristics are  particularly  suited  to  the  job  at 
hand.  Air  is  a very  good  dielectric  material, 
but  an  air-spaced  capacitor  does  not  have  a 
high  capacitance  since  the  dielectric  constant 
of  air  is  only  slightly  greater  than  one.  A 
group  of  other  commonly  used  dielectric  mate- 
ials  is  listed  in  figure  15. 

Certain  materials,  such  as  bakelite.  Incite, 
and  other  plastics  dissipate  considerable 
energy  when  used  as  capacitor  dielectrics. 


TABLE  OF  DIELECTRIC 

materials 

MATERIAL 

DIELECTRIC 
CONSTANT 
10  MC. 

POWER 
FACTOR 
tOMC.  1 

SOFTENING 

POINT 

FAHRENHEIT 

ANILINE -FORMALDEHYDE 
AE5IN 

3.4 

0.004 

260* 

BARIUM  TITANATE 

1200 

1.0 

— 

CASTOR  OIL 

4.67 

CELLULOSE  ACETATE 

3,7 

0.04 

180" 

CLASS,  WINDOW 

8-8 

POOR 

2000* 

CLASS.  PYREX 

4.5 

0.02 

KEL-F  FLUOROTHENE 

2.5 

0.6 

— 

METHYL-  METHACRYLATE  — 
LUCITE 

2.6 

0.007 

160® 

MICA 

5.4 

0.0003 

MYCALEX,  MYKROY 

7.0 

0.002 

650“ 

PHENOL- FORMALDEHYDE, 
LOW-LOSS  YELLOW 

5.0 

0.015 

270“ 

PHENOL- FORMALDEHYDE 
BLACK  BAKELITE 

5.5 

0.03 

350“ 

PORCELAIN 

7.0 

0.005 

2600* 

POLYETHYLENE  1 

2.25 

0.0003 

220“ 

POLYSTYRENE 

2.55 

0.0002 

175“ 

QUAHTE. FUSED 

4.2 

0.0002 

2600“ 

RUBBER.  HARD-EBONITE 

2.8 

0.007 

150“ 

STEATITE 

6. 1 

0.003 

2 700“ 

SULFUR 

3.6 

0.003 

236“ 

TEFLON 

2.  t 

0.02 

— 

TITANIUM  DIOXIDE 

100-175 

0.0006 

2700“ 

TRANSFORMER  OIL 

2.2 

0.003 

UREA  -FORMALDEHYDE 

5.0 

0.05 

260“ 

VINYL  RESINS 

4.0 

0.02 

200“ 

WOOD,  MAPLE 

4.4 

POOR 

FIGURE  15 


32  Direct  Current  Circuits 


THE  RADIO 


This  energy  loss  is  expressed  in  terms  of  the 
power  factor  of  the  capacitor,  which  repre- 
sents the  portion  of  the  input  volt-amperes 
lost  in  the  dielectric  material.  Other  materials 
including  air,  polystyrene  and  quartz  have  a 
very  low  power  factor. 

The  new  ceramic  dielectrics  such  as  stea- 
tite (talc)  and  titanium  dioxide  products  are 
especially  suited  for  high  frequency  and  high 
temperature  operation.  Ceramics  based  upon 
titanium  dioxide  have  an  unusually  high  di- 
electric constant  combined  with  a low  power 
factor.  The  temperature  coefficient  with  re- 
spect to  capacity  of  units  made  with  this 
material  depends  upon  the  mixture  of  oxides, 
and  coefficients  ranging  from  zero  to  over 
-700  parts  per  million  per  degree  Centigrade 
may  be  obtained  in  commercial  production. 

Mycalex  is  a composition  of  minute  mica 
particles  and  lead  borate  glass,  mixed  and 
fired  at  a relatively  low  temperature.  It  is  hard 
and  brittle,  but  can  be  drilled  or  machined 
when  water  is  used  as  the  cutting  lubricant. 

Mica  dielectric  capacitors  have  a very  low 
power  factor  and  extremely  high  voltage  break- 
down per  unit  of  thickness.  A mica  and  copper- 
foil  "sandwich”  is  formed  under  pressure  to 
obtain  the  desired  capacity  value.  The  effect 
of  temperature  upon  the  pressures  in  the 
"sandwich”  causes  the  capacity  of  the  usual 
mica  capacitor  to  have  large,  non-cyclic  vari- 
ations. If  the  copper  electrodes  ate  plated 
directly  upon  the  mica  sheets,  the  temperature 
coefficient  can  be  stablized  at  about  20  parts 
pet  million  pet  degree  Centigrade.  A process 
of  this  type  is  used  in  the  manufacture  of 
"silver  mica”  capacitors. 

Paper  dielectric  capacitors  consist  of  strips 
of  aluminum  foil  insulated  from  each  other  by 
a thin  layer  of  paper,  the  whole  assembly  being 
wrapped  in  a circular  bundle.  The  cost  of  such 
a capacitor  is  low,  the  capacity  is  high  in 
proportion  to  the  size  and  weight,  and  the 
power  factor  is  good.  The  life  of  such  a ca- 
pacitoris  dependent  upon  the  moisture  penetra- 
tion of  the  paper  dielectric  and  upon  the  ap- 
plied d-c  voltage. 

Air  dielectric  capacitors  are  used  in  trans- 
mitting and  receiving  circuits,  principally 
where  a variable  capacitor  of  high  tesetability 
is  required.  The  dielectric  strength  is  high, 
though  somewhat  less  at  radio  frequencies 
than  at  60  cycles.  In  addition,  corona  dis- 
charge at  high  frequencies  will  cause  ioniza- 
tion of  the  air  dielectric  causing  an  increase 
in  power  loss.  Dielectric  strength  may  be  in- 
creased by  increasing  the  air  pressure,  as  is 
done  in  hermetically  sealed  radar  units.  In 
some  units,  dry  nitrogen  gas  may  be  used  in 
place  of  air  to  provide  a higher  dielectric 
strength  than  that  of  air. 

Likewise,  the  dielectric  strength  of  an  "air” 


capacitor  may  be  increased  by  placing  the 

unit  in  a vacuum  chamber  to  prevent  ionization 
of  the  dielectric. 

The  temperature  coefficient  of  a variable 
air  dielectric  capacitor  varies  widely  and  is 
often  non-cyclic.  Such  things  as  differential 
expansion  of  various  parts  of  the  capacitor, 
changes  in  internal  stresses  and  different 

temperature  coefficients  of  various  parts  con- 
tribute to  these  variances. 

Dielectric  The  capacitance  of  a capacitor  is 
Constant  determined  by  the  thickness  and 
nature  of  the  dielectric  material 

between  plates.  Certain  materials  offer  a 

greater  capacitance  than  others,  depending 

upon  their  physical  makeup  and  chemical  con- 
stitution. This  property  is  expressed  by  a 
constant  K,  called  the  dielectric  constant. 
(K  = 1 fot  air.) 

Dielectric  If  the  charge  becomes  too  great 
Breakdown  fot  a given  thickness  of  a certain 
dielectric,  the  capacitor  will  break 
down,  i.  e.,  the  dielectric  will  puncture.  It  is 
fot  this  reason  that  capacitors  are  rated  in  the 
manner  of  the  amount  of  voltage  they  will 
safely  withstand  as  well  as  the  capacitance 
in  microfarads.  This  rating  is  commonly  ex- 
pressed as  the  d-c  working  voltage. 


Calculation  of  The  capacitance  of  two  parallel 
Copacitance  plates  is  given  with  good  accu- 
racy by  the  following  formula: 


CAPACITANCE  IN  MICRO-MICROFARADS 

Figure  16 

Through  tho  use  of  this  chart  It  Is  possible  to 
determine  the  roquirod  plats  dlamstsr  (with  the 
the  necessary  spacing  sstabllshsd  by  peak 
voltage  considerations)  for  a circular-plate 
neutralising  capacitor.  The  capacitance  given 
Is  for  a dielectric  of  air  and  the  spacing  given 
Is  between  ad/acent  faces  af  the  two  plates. 


HANDBOOK 


Capacitive  Circuits  33 


V 1 t 1 

PARALLEL  CAPACITORS  SERIES  CAPACITORS 


cc. 

-C3  7 

=c.  1 

c 

=C2  = 

-C4  ^ 

::tC6  1 

CAPACITORS  IN  SERIES-PARALLEL 


Figure  17 

CAPACITORS  IN  SERIES,  PARALLEL, 
AND  SERIES-PARALLEL 


A 

C =0.2248  X K X— , 
t 

where  C = capacitance  in  micro-microfarads, 
K = dielectric  constant  of  spacing 
material, 

A = area  of  dielectric  in  square  inches, 
t = thickness  of  dielectric  in  inches. 

This  formula  indicates  that  the  capacitance 
is  directly  proportional  to  the  area  of  the 
plates  and  inversely  proportional  to  the  thick- 
ness of  the  dielectric  (spacing  between  the 
plates).  This  simply  means  that  when  the  area 
of  the  plate  is  doubled,  the  spacing  between 
plates  remaining  constant,  the  capacitance 
will  be  doubled.  Also,  if  the  area  of  the  plates 
remains  constant,  and  the  plate  spacing  is 
doubled,  the  capacitance  will  be  reduced  to 
half. 

The  above  equation  also  shows  that  capaci- 
tance is  directly  proportional  to  the  dielectric 
constant  of  the  spacing  material.  An  air-spaced 
capacitor  that  has  a capacitance  of  100  /tjufd. 
in  air  would  have  a capacitance  of  467  fififd. 
when  immersed  in  castor  oil,  because  the  di- 
electric constant  of  castor  oil  is  4.67  times  as 
great  as  the  dielectric  constant  of  ait. 

Where  the  area  of  the  plates  is  definitely 
set,  when  it  is  desired  to  know  the  spacing 
needed  to  secure  a required  capacitance, 

A X 0.2248  X K 

t 

C 

where  all  units  are  expressed  just  as  in  the 
preceding  formula.  This  formula  is  not  con- 
fined to  capacitors  having  only  square  or 


rectangular  plates,  but  also  applies  when  the 
plates  ate  circular  in  shape.  The  only  change 
will  be  the  calculation  of  the  area  of  such 
circular  plates;  this  area  can  be  computed  by 
squaring  the  radius  of  the  plate,  then  multiply- 
ing by  3.1416,  or  "pi.”  Expressed  as  an 
equation: 

A = 3.1416  xr" 

where  r = radius  in  inches 

The  capacitance  of  a multi-plate  capacitor 
can  be  calculated  by  taking  the  capacitance  of 
one  section  and  multiplying  this  by  the  number 
of  dielectric  spaces.  In  such  cases,  however, 
the  formula  gives  no  consideration  to  the 
effects  of  edge  capacitance;  so  the  capaci- 
tance as  calculated  will  not  be  entirely  ac- 
curate. These  additional  capacitances  will  be 
but  a small  part  of  the  effective  total  capaci- 
tance, particularly  when  the  plates  are  reason- 
ably large  and  thin,  and  the  final  result  will, 
therefore,  be  within  practical  limits  of  accuracy. 


Capacitors  in  Equations  for  calculating  ca- 
Parallel  and  pacitances  of  capacitors  in  par- 
in  Series  allel  connections  are  the  same 

as  those  for  resistors  in  series. 


C = Cj  -f.  Cj  + , . » -I-  Cn 

Capacitors  in  series  connection  ate  cal- 
culated in  the  same  manner  as  ate  resistors  in 
parallel  connection. 

The  formulas  are  repeated:  (1)  For  two  or 
more  capacitors  of  unequal  capacitance  in 
series: 


C =- 


1 1 1 

— 4- + — 

C,  c,  c, 

1111 

or  — = — H -I 

C C,  C,  C3 

(2)  Two  capacitors  of  unequal  capacitance  in 
series: 

C,  x C, 

C = 

Cj  + Cj 

(3)  Three  capacitors  of  equal  capacitance  in 


C, 

C = where  Cj  is  the  common  capacitance. 

3 

(4)  Three  or  more  capacitors  of  equal  capac- 
itance in  series. 


Value  of  common  capacitance 

C = ^ ^ ^ 

Number  of  capacitors  in  series 

(5)  Six  capacitors  in  series  parallel: 


34  Direct  Current  Circuits 


THE  RADIO 


C.  C,  C3  c,  c,  c. 


Capacitors  in  A'C  When  a capacitor  is  con- 
and  D-C  Circuits  nected  into  a direct-cur- 
rent  circuit,  it  will  block 
the  d.c.,  or  stop  the  flow  of  current.  Beyond 
the  initial  movement  of  electrons  during  the 
period  when  the  capacitor  is  being  charged, 
there  will  be  no  flow  of  current  because  the 
circuit  is  effectively  broken  by  the  dielectric 
of  the  capacitor. 

Strictly  speaking,  a very  small  current  may 
actually  flow  because  the  dielectric  of  the 
capacitor  may  not  be  a perfect  insulator.  This 
minute  current  flow  is  the  leakage  current 
previously  referred  to  and  is  dependent  upon 
the  internal  d-c  resistance  of  the  capacitor. 
This  leakage  current  is  usually  quite  notice- 
able in  most  types  of  electrolytic  capacitors. 

When  an  alternating  current  is  applied  to 
a capacitor,  the  capacitor  will  charge  and  dis- 
charge a certain  number  of  times  per  second 
in  accordance  with  the  frequency  of  the  alter- 
nating voltage.  The  electron  flow  in  the  charge 
and  discharge  of  a capacitor  when  an  a-c 
potential  is  applied  constitutes  an  alternating 
current,  in  effect.  It  is  for  this  reason  that  a 
capacitor  will  pass  an  alternating  current  yet 
offer  practically  infinite  opposition  to  a direct 
current.  These  two  properties  ate  repeatedly 
in  evidence  in  a radio  circuit. 

Voltage  Rating  Any  good  paper  dielectric 
of  Capacitors  filter  capacitor  has  such  a 
in  Series  high  internal  resistance  (in- 

dicating a good  dielectric) 
that  the  exact  resistance  will  vary  consider- 
ably from  capacitor  to  capacitor  even  though 
they  are  made  by  the  same  manufacturer  and 
ate  of  the  same  rating.  Thus,  when  1000  volts 
d.c.  is  connected  across  two  l-pfd.  500-volt 
capacitors  in  series,  the  chances  ace  that  the 
voltage  will  divide  unevenly  and  one  capacitor 
will  receive  more  than  500  volts  and  the  other 
less  than  500  volts. 

Voltage  Equalizing  By  connecting  a half- 
Resistors  megohm  1-watt  carbon  re- 

sistor across  each  capac- 
itor, the  voltage  will  be  equalized  because  the 
resistors  act  as  a voltage  divider,  and  the 
internal  resistances  of  the  capacitors  are  so 
much  higher  (many  megohms)  that  they  have 
but  little  effect  in  disturbing  the  voltage  di- 
vider balance. 

Carbon  resistors  of  the  inexpensive  type 
are  not  particularly  accurate  (not  being  de- 
signed for  precision  service);  therefore  it  is 


Figure  18 

SHOWING  THE  USE  OF  VOLTAGE  EQUAL- 
IZING RESISTORS  ACROSS  CAPACITORS 
CONNECTED  IN  SERIES 


advisable  to  check  several  on  an  accurate 
ohmmeter  to  find  two  that  are  as  close  as 
possible  in  resistance.  The  exact  resistance 
is  unimportant,  just  so  it  is  the  same  for  the 
two  resistors  used. 

Capacitors  in  When  two  capacitors  are  con- 
SeriesonA.C.  nected  in  series,  alternating 
voltage  pays  no  heed  to  the 
relatively  high  internal  resistance  of  each 
capacitor,  but  divides  across  the  capacitors 
in  inverse  proportion  to  the  capacitance.  Be- 
cause, in  addition  to  the  d.c.  across  a capac- 
itor in  a filter  or  audio  amplifier  circuit  there 
is  usually  an  a-c  or  a-f  voltage  component,  it 
is  inadvisable  to  series- connect  capacitors 
of  unequal  capacitance  even  if  dividers  are 
provided  to  keep  the  d.c.  within  the  ratings  of 
the  individual  capacitors. 

For  instance,  if  a 500-volt  1-ftfd.  capacitor 
is  used  in  series  with  a 4-ftfd.  500-volt  capac- 
itor across  a 250-volt  a-c  supply,  the  1-pfd. 
capacitor  will  have  200  volts  a.c.  across  it 
and  the  4-fifd.  capacitor  only  50  volts.  An 
equalizing  divider  to  do  any  good  in  this  case 
would  have  to  be  of  very  low  resistance  be- 
cause of  the  comparatively  low  impedance  of 
the  capacitors  to  a.c.  Such  a divider  would 
draw  excessive  current  and  be  impracticable. 

The  safest  rule  to  follow  is  to  use  only 
capacitors  of  the  same  capacitance  and  volt- 
age rating  and  to  install  matched  high  resist- 
ance proportioning  resistors  across  the  various 
capacitors  to  equalize  the  d-c  voltage  drop 
across  each  capacitor.  This  holds  regardless 
of  how  many  capacitors  are  series-connected. 

Electrolytic  Electrolytic  capacitors  use  a very 
Capacitors  thin  film  of  oxide  as  the  dielec- 
tric, and  are  polarized;  that  is, 
they  have  a positive  and  a negative  terminal 
which  must  be  properly  connected  in  a circuit; 
otherwise,  the  oxide  will  break  down  and  the 
capacitor  will  overheat.  The  unit  then  will  no 
longer  be  of  service.  When  electrolytic  capac- 
itors are  connected  in  series,  the  positive  ter- 
minal is  always  connected  to  the  positive  lead 
of  the  power  supply;  the  negative  terminal  of 


HANDBOOK 


Magnetism  35 


the  capacitor  connects  to  the  positive  terminal 
of  the  next  capacitor  in  the  series  combination. 
The  method  of  connection  for  electrolytic  ca- 
pacitors in  series  is  shown  in  figure  18.  Elec- 
trolytic capacitors  have  very  low  cost  per 
microfarad  of  capacity,  but  also  have  a large 
power  factor  and  high  leakage;  both  dependent 
upon  applied  voltage,  temperature  and  the  age 
of  the  capacitor.  The  modern  electrolytic  ca- 
pacitor uses  a dry  paste  electrolyte  embedded 
in  a gauze  or  paper  dielectric.  Aluminium  foil 
and  the  dielectric  are  wrapped  in  a circular 
bundle  and  are  mounted  in  a cardboard  or  metal 
box.  Etched  electrodes  may  be  employed  to 
increase  the  effective  anode  area,  and  the 
total  capacity  of  the  unit. 

The  capacity  of  an  electrolytic  capacitor  is 
affected  by  the  applied  voltage,  the  usage  of 
the  capacitor,  and  the  temperature  and  humidity 
of  the  environment.  The  capacity  usually  drops 
with  the  aging  of  the  unit.  The  leakage  current 
and  power  factor  increase  with  age.  At  high 
frequencies  the  power  factor  becomes  so  poor 
that  the  electrolytic  capacitor  acts  as  a series 
resistance  rather  than  as  a capacity. 

2-4  MagneMsm 

and  Electromagnetism 

The  common  bar  or  horseshoe  magnet  is 
familiar  to  most  people.  The  magnetic  field 
which  surrounds  it  causes  the  magnet  to  at- 
tract other  magnetic  materials,  such  as  iron 
nails  or  tacks.  Exactly  the  same  kind  of  mag- 
netic field  is  set  up  around  any  conductor 
carrying  a current,  but  the  field  exists  only 
while  the  current  is  flowing. 

Magnetic  Fields  Before  a potential,  or  volt- 
age, is  applied  to  a con- 
ductor there  is  no  external  field,  because  there 
is  no  general  movement  of  the  electrons  in 
one  direction.  However,  the  electrons  do  pro- 
gressively move  along  the  conductor  when  an 
e.m.f.  is  applied,  the  direction  of  motion  de- 
pending upon  the  polarity  of  the  e.m.f.  Since 
each  electron  has  an  electric  field  about  it,  the 
flow  of  electrons  causes  these  fields  to  build 
up  into  a resultant  external  field  which  acts  in 
a plane  at  right  angles  to  the  direction  in 
which  the  current  is  flowing.  This  field  is 
known  as  the  magnetic  field. 

The  magnetic  field  around  a current-carrying 
conductor  is  illustrated  in  figure  19.  The 
direction  of  this  magnetic  field  depends  en- 
tirely upon  the  direction  of  electron  drift  or 
current  flow  in  the  conductor.  When  the  flow 
is  toward  the  observer,  the  field  about  the 
conductor  is  clockwise;  when  the  flow  is  away 
from  the  observer,  the  field  is  counter-clock- 
wise. This  is  easily  remembered  if  the  left 
hand  is  clenched,  with  the  thumb  outstretched 


Figure  19 

UEFT-HANO  RULE 

Showing  tho  diroction  of  tho  magnotic  linos  of 
foreo  produeod  around  a conductor  carrying  an 
oloetric  currant. 


and  pointing  in  the  direction  of  electron  flow. 
The  fingers  then  indicate  the  direction  of  the 
magnetic  field  around  the  conductor. 

Each  electron  adds  its  field  to  the  total  ex- 
ternal magnetic  field,  so  that  the  greater  the 
number  of  electrons  moving  along  the  con- 
ductor, the  stronger  will  be  the  resulting  field. 

One  of  the  fundamental  laws  of  magnetism 
is  that  like  poles  repel  one  another  and  unlike 
poles  attract  one  another.  This  is  true  of  cur- 
rent-carrying conductors  as  well  as  of  perman- 
ent magnets.  Thus,  if  two  conductors  are  placed 
side  by  side  and  the  current  in  each  is  flowing 
in  the  same  direction,  the  magnetic  fields  will 
also  be  in  the  same  direction  and  will  combine 
to  form  a larger  and  stronger  field.  If  the  cur- 
rent flow  in  adjacent  conductors  is  in  opposite 
directions,  the  magnetic  fields  oppose  each 
other  and  tend  to  cancel. 

The  magnetic  field  around  a conductor  may 
be  considerably  increased  in  strength  by  wind- 
ing the  wire  into  a coil.  The  field  around  each 
wire  then  combines  with  those  of  the  adjacent 
turns  to  form  a total  field  through  the  coil 
which  is  concentrated  along  the  axis  of  the 
coil  and  behaves  externally  in  a way  similar 
to  the  field  of  a bar  magnet. 

If  the  left  hand  is  held  so  that  the  thumb 
is  outstretched  and  parallel  to  the  axis  of  a 
coil,  with  the  fingers  curled  to  indicate  the 
direction  of  electron  flow  around  the  turns  of 
the  coil,  the  thumb  then  points  in  the  direc- 
tion of  the  north  pole  of  the  magnetic  field. 

Th®  Magnetic  In  the  magnetic  circuit,  the 
units  which  correspond  to  cur- 
^ rent,  voltage,  and  resistance 

in  the  electrical  circuit  are  flux,  magneto- 
motive force,  and  reluctance. 

Flux,  Flux  As  a current  is  made  up  of  a drift 
Density  of  electrons,  so  is  a magnetic 

field  made  up  of  lines  of  force,  and 
the  total  number  of  lines  of  force  in  a given 
magnetic  circuit  is  termed  the  flux.  The  flux 
depends  upon  the  material,  cross  section,  and 
length  of  the  magnetic  circuit,  and  it  varies 
directly  as  the  current  flowing  in  the  circuit. 


36  Direct  Current  Circuits 


THE  RADIO 


The  unit  of  flux  is  the  maxwell,  and  the  sym- 
bol is  the  Greek  letter  (phi). 

Flux  density  is  the  number  of  lines  of  force 
per  unit  area.  It  is  expressed  in  gauss  if  the 
unit  of  area  is  the  square  centimeter  (1  gauss 
= 1 line  of  force  per  square  centimeter),  or 
in  lines  per  square  inch.  The  symbol  for  flux 
density  is  B if  it  is  expressed  in  gausses,  or 
B if  expressed  in  lines  per  square  inch. 

Magnetomotive  The  force  which  produces  a 
Force  flux  in  a magnetic  circuit 

is  called  magnetomotive  force. 
It  is  abbreviated  m.m.f.  and  is  designated  by 
the  letter  F.  The  unit  of  magnetomotive  force 
is  the  gilbert,  which  is  equivalent  to  1.26  X NI, 
where  N is  the  number  of  turns  and  I is  the 
current  flowing  in  the  circuit  in  amperes. 

The  m.m.f.  necessary  to  produce  a given 
flux  density  is  stated  in  gilberts  per  centi- 
meter (oersteds)  (H),  or  in  ampere-turns  per 
inch  (H). 

Reluctance  Magnetic  reluctance  corresponds 
to  electrical  resistance,  and  is 
the  property  of  a material  that  opposes  the 
creation  of  a magnetic  flux  in  the  material. 
It  is  expressed  in  rels,  and  the  symbol  is  the 
letter  R.  A material  has  a reluctance  of  1 rel 
when  an  m.m.f.  of  1 ampere-turn  (NI)  generates 
a flux  of  1 line  of  force  in  it.  Combinations 
of  reluctances  are  treated  the  same  as  re- 
sistances in  finding  the  total  effective  reluc- 
tance. The  specific  reluctance  of  any  sub- 
stance is  its  reluctance  pet  unit  volume. 

Except  for  iron  and  its  alloys,  most  common 
materials  have  a specific  reluctance  very 
nearly  the  same  as  that  of  a vacuum,  which, 
for  all  practical  purposes,  may  be  considered 
the  same  as  the  specific  reluctance  of  air. 

Ohm's  Low  for  The  relations  between  flux. 
Magnetic  Circuits  magnetomotive  force,  and 
reluctance  ate  exactly  the 
same  as  the  relations  between  current,  volt- 
age, and  resistance  in  the  electrical  circuit. 
These  can  be  stated  as  follows; 

F F 

<A=—  R=—  F = (AR 

^ R <f> 

where  cf>  = flux,  F = m.m.f.,  and  R = reluctance. 

Permeability  Permeability  expresses  the  ease 

with  which  a magnetic  field  may 

be  set  up  in  a material  as  compared  with  the 
effort  requited  in  the  case  of  ait.  Iron,  for  ex- 
ample, has  a permeability  of  around  2000 
times  that  of  air,  which  means  that  a given 
amount  of  magnetizing  effect  produced  in  an 
iron  core  by  a current  flowing  through  a coil 
of  wire  will  produce  2000  times  the  flux  density 
that  the  same  magnetizing  effect  would  pro- 


duce in  air.  It  may  be  expressed  by  the  ratio 

B/H  or  B/H.  In  other  words, 

B B 


where  p is  the  premeability,  B is  the  flux 
density  in  gausses,  B is  the  flux  density  in 
lines  pet  square  inch,  H is  the  m.m.f.  in 
gilberts  pet  centimeter  (oersteds),  and  H is 
the  m.m.f.  in  ampere-turns  pet  inch.  These 
relations  may  also  be  stated  as  follows: 

B B 

H = — or  H = — , and  B = Hp  or  B = Hfr 

F F 

It  can  be  seen  from  the  foregoing  that  per- 
meability is  inversely  proportional  to  the 
specific  reluctance  of  a material. 

Saturation  Permeability  is  similar  to  electric 
conductivity.  There  is,  however, 
one  important  difference:  the  permeability  of 
magnetic  materials  is  not  independent  of  the 
magnetic  current  (flux)  flowing  through  it, 
although  electrical  conductivity  is  substan- 
tially independent  of  the  electric  current  in  a 
wire.  When  the  flux  density  of  a magnetic 
conductor  has  been  increased  to  the  saturation 
point,  a further  increase  in  the  magnetizing 
force  will  not  produce  a corresponding  in- 
crease in  flux  density. 

Calculations  To  simplify  magnetic  circuit 
calculations,  a magnetization 
curve  may  be  drawn  for  a given  unit  of  ma- 
terial. Such  a curve  is  termed  a B-H  curve,  and 
may  be  determined  by  experiment.  When  the 
current  in  an  iron  cote  coil  is  first  applied, 
the  relation  between  the  winding  current  and 
the  core  flux  is  shown  at  A-B  in  figure  20.  If 
the  current  is  then  reduced  to  zero,  reversed, 
brought  back  again  to  zero  and  reversed  to  the 


TYPICAL  HYSTERESIS  LOOP 
(B-H  CURVE  = A-B) 

Showing  relationship  between  the  current  in  the 
winding  of  an  iron  core  inductor  and  the  core 
flus.  A direct  current  flowing  through  the  Induce 
tance  brings  the  magnetic  state  of  the  core  to 
some  point  on  the  hysteresis  loop,  such  as  C. 


HANDBOOK 


Inductance  37 


original  direction,  the  flux  passes  through  a 
typical  hysteresis  loop  as  shown. 

Residual  Magnetism;  The  magnetism  remaining 
Retentivity  in  a material  after  the 

magnetizing  force  is  re- 
moved is  called  residual  magnetism.  Reten- 
tivity is  the  property  which  causes  a magnetic 
material  to  have  residual  magnetism  after 
having  been  magnetized. 

Hysteresis;  Hysteresis  is  the  character- 

Coercive  Force  istic  of  a magnetic  system 
which  causes  a loss  of  power 
due  to  the  fact  that  a negative  magnetizing 
force  must  be  applied  to  reduce  the  residual 
magnetism  to  zero.  This  negative  force  is 
termed  coercive  force.  By  **negative**  mag- 
netizing force  is  meant  one  which  is  of  the 
opposite  polarity  with  respect  to  the  original 
magnetizing  force.  Hysteresis  loss  is  apparent 
in  transformers  and  chokes  by  the  heating  of 
the  core. 

Inductance  If  the  switch  shown  in  figure  19 
is  opened  and  closed,  a pulsating 
direct  current  will  be  produced.  When  it  is 
first  closed,  the  current  does  not  instanta- 
neously rise  to  its  maximum  value,  but  builds 
up  to  it.  While  it  is  building  up,  the  magnetic 
field  is  expanding  around  the  conductor.  Of 
course,  this  happens  in  a small  fraction  of  a 
second.  If  the  switch  is  then  opened,  the  cur- 
rent stops  and  the  magnetic  field  contracts 
quickly.  This  expanding  and  contracting  field 
will  induce  a current  in  any  other  conductor 
that  is  part  of  a continuous  circuit  which  it 
cuts.  Such  a field  can  be  obtained  in  the  way 
just  mentioned  by  means  of  a vibrator  inter- 
ruptor,  or  by  applying  a.c.  to  the  circuit  in 
place  of  the  battery.  Varying  the  resistance  of 
the  circuit  will  also  produce  the  same  effect. 
This  inducing  of  a current  in  a conductor  due 
to  a varying  current  in  another  conductor  not 
in  acutal  contact  is  called  electromagnetic  in- 
duction. 


Self-inductance  If  an  alternating  current  flows 
through  a coil  the  varying 
magnetic  field  around  each  turn  cuts  itself  and 
the  adjacent  turn  and  induces  a voltage  in  the 
coil  of  opposite  polarity  to  the  applied  e.m.f. 
The  amount  of  induced  voltage  depends  upon 
the  number  of  turns  in  the  coil,  the  current 
flowing  in  the  coil,  and  the  number  of  lines 
of  force  threading  the  coil.  The  voltage  so 
induced  is  known  as  a counter-e.m.f.  or  back- 
e.m.f.,  and  the  effect  is  termed  self-induction. 
When  the  applied  voltage  is  building  up,  the 
counter-e.m.f.  opposes  the  rise;  when  the  ap- 
plied voltage  is  decreasing,  the  counter-e.m.f. 
is  of  the  same  polarity  and  tends  to  maintain 


the  current.  Thus,  it  can  be  seen  that  self- 
induction  tends  to  prevent  any  change  in  the 
current  in  the  circuit. 

The  storage  of  energy  in  a magnetic  field 
is  expressed  in  joules  and  is  equal  to  (LI^)/2. 
(A  joule  is  equal  to  1 watt-second.  L is  de- 
fined immediately  following.) 

The  Unit  of  Inductance  is  usually  denoted  by 
Inductance;  the  letter  L,  and  is  expressed  in 
The  Henry  henrys.  A coil  has  an  inductance 
of  1 henry  when  a voltage  of  1 
volt  is  induced  by  a current  change  of  1 am- 
pere per  second.  The  henry,  while  commonly 
used  in  audio  frequency  circuits,  is  too  large 
for  reference  to  inductance  coils,  such  as 
those  used  in  radio  frequency  circuits;  milli- 
henry or  microhenry  is  more  commonly  used, 
in  the  following  manner: 

1 henry  = 1,000  millihenry  s,  or  10^  milli- 
henrys. 

1 millihenry  1/1,000  of  a henry,  .001  henry, 
or  lO""^  henry. 

1 wtcro^enry « 1/1,000,000  of  a henry,  or 
.000001  henry,  or  10"®  henry. 

1 microhenry  ^ \/\y000  of  a millihenry,  .001 
or  10“*  millihenrys. 

1,000  mtcrohenrys  « 1 millihenry. 


Mutual  Inductance  When  one  coil  is  near  an- 
other, a varying  current  in 
one  will  produce  a varying  magnetic  field 
which  cuts  the  turns  of  the  other  coil,  inducing 
a current  in  it.  This  induced  current  is  also 
varying,  and  will  therefore  induce  another  cur- 
rent in  the  first  coil.  This  reaction  between 
two  coupled  circuits  is  called  mutual  induction, 
and  can  be  calculated  and  expressed  in  henrys. 
The  symbol  for  mutual  inductance  is  M.  Two 
circuits  thus  joined  are  said  to  be  inductively 
coupled. 

The  magnitude  of  the  mutual  inductance  de- 
pends upon  the  shape  and  size  of  the  two  cir- 
cuits, their  positions  and  distances  apart,  and 
the  premeability  of  the  medium.  The  extent  to 


p5Wj 


Figure  21 

MUTUAL  INDUCTANCE 

T/te  quanfity  M repr^s^nts  f/ie  mutual  Induetanca 
hetwan  tha  two  colls  Lj  and  L2. 


38  Direct  Current  Circuits 


THE  RADIO 


INDUCTANCE  OF 
SINGLE-LAYER 
SOLENOID  COILS 


R2  N2 
9R+10L 


MICROHENRIES 


WHERE  : R = RADIUS  OF  COIL  TO  CENTER  OF  WIRE 
L = LENGTH  OF  COIL 
N = NUMBER  OF  TURNS 


Figure  22 
FORMULA  FOR 
CALCULATING  INDUCTANCE 

Through  the  use  of  the  equation  and  the  sketch 
shown  above  the  Inductance  of  single-layer 
solenoid  coils  can  be  calculated  with  an  ac- 
curacy of  about  one  per  cent  for  the  types  of 
colls  normally  used  In  the  h-f  and  v-h-f  range. 


which  two  inductors  are  coupled  is  expressed 
by  a relation  known  as  coefficient  of  coupling. 
This  is  the  ratio  of  the  mutual  inductance  ac- 
tually present  to  the  maximum  possible  value. 

The  formula  for  mutual  inductance  is  L = 
L,  + L,  + 2M  when  the  coils  ate  poled  so  that 
their  fields  add.  When  they  are  poled  so  that 
their  fields  buck,  then  L = Lj  + L,  - 2M 
(figure  21). 

Inductors  in  Inductors  in  parallel  are  com- 
Porallel  bined  exactly  as  ate  resistors  in 
parallel,  provided  that  they  are 
fat  enough  apart  so  that  the  mutual  inductance 
is  entirely  negligible. 

Inductors  in  Inductors  in  series  are  additive, 
Series  just  as  ate  resistors  in  series, 

again  provided  that  no  mutual 
inductance  exists.  In  this  case,  the  total  in- 
ductance L is: 

L = L,  + Lj  + etc. 

Where  mutual  inductance  does  exist: 

L =L,  +Lj  -I-  2M, 
where  M is  the  mutual  inductance. 

This  latter  expression  assumes  that  the 
coils  are  connected  in  such  a way  that  all  flux 
linkages  are  in  the  same  direction,  i.e.,  ad- 
ditive. If  this  is  not  the  case  and  the  mutual 
linkages  subtract  from  the  self-linkages,  the 
following  fotmula  holds: 

L = Li  + Lj  - 2M, 
where  M is  the  mutual  inductance. 


Core  Material  Ordinary  magnetic  cores  can- 
not be  used  for  tadio  frequen- 
cies because  the  eddy  current  and  hysteresis 
losses  in  the  cote  material  becomes  enormous 


as  the  frequency  is  increased.  The  principal 
use  for  conventional  magnetic  cotes  is  in  the 
audio- frequency  range  below  approximately 
15,000  cycles,  whereas  at  very  low  frequencies 
(50  to  60  cycles)  their  use  is  mandatory  if 
an  appreciable  value  of  inductance  is  desired. 

An  air  core  inductor  of  only  1 henry  in- 
ductance would  be  quite  latge  in  size,  yet 
values  as  high  as  500  henrys  are  commonly 
available  in  small  iron  cote  chokes.  The  in- 
ductance of  a coil  with  a magnetic  core  will 
vary  with  the  amount  of  current  (both  a-c  and 
d-c)  which  passes  through  the  coil.  For  this 
reason,  iron  core  chokes  that  are  used  in  power 
supplies  have  a certain  inductance  rating  at  a 
predetermined  value  of  d-c. 

The  premeability  of  ait  does  not  change 
with  flux  density;  so  the  inductance  of  iron 
core  coils  often  is  made  less  dependent  upon 
flux  density  by  making  part  of  the  magnetic 
path  ait,  instead  of  utilizing  a closed  loop  of 
iron.  This  incorporation  of  an  air  gap  is  nec- 
essary in  many  applications  of  iron  core  coils, 
particularly  where  the  coil  carries  a consider- 
able d-c  component.  Because  the  permeability 
of  air  is  so  much  lower  than  that  of  iron,  the 
ait  gap  need  comprise  only  a small  fraction  of 
the  magnetic  circuit  in  otdet  to  provide  a sub- 
stantial proportion  of  the  total  reluctance. 

Iron  Cored  Inductors  Iron-core  inductors  may 
at  Rodio  Frequencies  be  used  at  radio  frequen- 
cies if  the  iron  is  in  a 
very  finely  divided  form,  as  in  the  case  of  the 
powdered  iron  cores  used  in  some  types  of  r-f 
coils  and  i-f  transformers.  These  cores  are 
made  of  extremely  small  particles  of  iron.  The 
particles  are  treated  with  an  insulating  mater- 
ial so  that  each  particle  will  be  insulated  from 
the  others,  and  the  treated  powder  is  molded 
with  a binder  into  cotes.  Eddy  current  losses 
ate  greatly  reduced,  with  the  result  that  these 
special  iron  cotes  ate  entirely  practical  in  cir- 
cuits which  operate  up  to  100  Me.  in  frequency. 


2-5  RC  and  RL  Transients 

A voltage  divider  may  be  constructed  as 
shown  in  figure  23.  Kirchhoff’s  and  Ohm’s 
Laws  hold  for  such  a divider.  This  circuit  is 
known  as  an  RC  circuit. 


Time  Constant-  When  switch  S in  figure  23  is 
RC  and  RL  placed  in  position  1,  a volt- 
Circuits  meter  across  capacitor  C will 

indicate  the  manner  in  which 
the  capacitor  will  become  charged  through  the 
resistor  R from  battery  B.  If  relatively  latge 
values  are  used  for  R and  C,  and  if  a v-t  volt- 
meter which  draws  negligible  current  is  used 


HANDBOOK 


Time  Constant  39 


Figure  23 

TIME  CONSTANT  OF  AN  R-C  CIRCUIT 

Shown  at  (A)  la  tha  circuit  upon  which  Is  based 
the  curves  of  (B)  and  (C).  (B)  shows  the  rate  at 
which  capacitor  C will  charge  from  the  Instant 
at  which  switch  S Is  placed  In  position  h (C) 
shows  the  discharge  curve  of  capacitor  C from 
the  Instant  at  which  switch  S is  placed  In 
position  3. 


to  measure  the  voltage  e,  the  rate  of  charge  of 
the  capacitor  may  actually  be  plotted  with  the 
aid  of  a stop  watch. 

Voltage  Gradient  It  will  be  found  that  the  volt- 
age e will  begin  to  rise 
rapidly  from  zero  the  instant  the  switch  is 
closed.  Then,  as  the  capacitor  begins  to 
charge,  the  rate  of  change  of  voltage  across 
the  capacitor  will  be  found  to  decrease,  the 
charging  taking  place  more  and  more  slowly 
as  the  capacitor  voltage  e approaches  the  bat- 
tery voltage  E.  Actually,  it  will  be  found  that 
in  any  given  interval  a constant  percentage  of 
the  remaining  difference  between  e and  E 
will  be  delivered  to  the  capacitor  as  an  in- 
crease in  voltage.  A voltage  which  changes  in 
this  manner  is  said  to  increase  logarithmically, 
or  is  said  to  follow  an  exponential  curve. 

Time  Constant  A mathematical  analysis  of 

the  charging  of  a capacitor  in 
this  manner  would  show  that  the  relationship 
between  the  battery  voltage  E and  the  voltage 
across  the  capacitor  e could  be  expressed  in 
the  following  manner: 

e = E (1 

where  e,E,R,  and  C have  the  values  discussed 
above,  e = 2.716  (the  base  of  Naperian  or 
natural  logarithms),  and  t represents  the  time 
which  has  elapsed  since  the  closing  of  the 
switch.  With  t expressed  in  seconds,  R and  C 


Figure  24 

TYPICAL  INDUCTANCES 

The  large  inductance  Is  o 1000-watt  transmitting  coif.  To  the  right  and  left  of  this  coil  are  small  r-f 
chokes.  Several  varieties  of  low  power  capability  colls  ere  shown  below,  along  with  various  types  of  r-f 
chokes  intended  for  high-freguency  operation. 


40  Direct  Current  Circuits 


Figure  25 

TIME  CONSTANT  OF  AN  R-L  CIRCUIT 

Note  that  the  time  constant  /or  the  increase  in 
current  through  on  R-L  c/rcoi#  is  identical  to 
the  rate  of  increase  in  voltage  across  the 
capacitor  In  an  R-C  circuit. 


may  be  expressed  in  farads  and  ohms,  or  R 
and  C may  be  expressed  in  microfarads  and 
megohms.  The  product  RC  is  called  the  time 
constant  of  the  circuit,  and  is  expressed  in 
seconds.  As  an  example,  if  R is  one  megohm 
and  C is  one  microfarad,  the  time  constant 
RC  will  be  equal  to  the  product  of  the  two, 
or  one  second. 

When  the  elapsed  time  t is  equal  to  the 
time  constant  of  the  RC  network  under  con- 
sideration, the  exponent  of  e becomes  -1. 
Now  is  equal  to  l/e,  or  1/2.716,  which 
is  0.368.  The  quantity  (1  - 0.368)  then  is  equal 
to  0.632.  Expressed  as  percentage,  the  above 


means  that  the  voltage  across  the  capaci- 
tor will  have  increased  to  63.2  per  cent  of 
the  battery  voltage  in  an  interval  equal  to  the 
time  constant  or  RC  product  of  the  circuit. 
Then,  during  the  next  period  equal  to  the  time 
constant  of  the  RC  combination,  the  voltage 
across  the  capacitor  will  have  risen  to  63.2 
per  cent  of  the  remaining  difference  in  voltage, 
or  86.5  per  cent  of  the  applied  voltage  E. 

RL  Circuit  In  the  case  of  a series  combination 
of  a resistor  and  an  inductor,  as 
shown  in  figure  25,  the  current  through  the 
combination  follows  a very  similar  law  to  that 
given  above  for  the  voltage  appearing  across 
the  capacitor  in  an  RC  series  circuit.  The 
equation  for  the  current  through  the  combina- 
tion is: 

E 

,■=_  (1  _f-tR/L) 

R 

where  i represents  the  current  at  any  instant 
through  the  series  circuit,  E represents  the 
applied  voltage,  and  R reptesents  the  total 
resistance  of  the  resistor  and  the  d-c  resist- 
ance of  the  inductor  in  series.  Thus  the  time 
constant  of  the  RL  circuit  is  L/R,  with  R ex- 
pressed in  ohms  and  L expressed  in  henrys. 

VoUoge  Decay  When  the  switch  in  figure  23  is 
moved  to  position  3 after  the 
capacitor  has  been  charged,  the  capacitor  volt- 
age will  drop  in  the  manner  shown  in  figure 
23-C.  In  this  case  the  voltage  across  the  ca- 
pacitor will  decrease  to  36.8  pet  cent  of  the 
initial  voltage  (will  make  63.2  per  cent  of  the 
total  drop)  in  a period  of  time  equal  to  the 
time  constant  of  the  RC  circuit. 


TYPICAL  IRON-CORE  INDUCTANCES 

At  the  right  is  an  upright  mounting  filter  choke  intended  for  use  in  low  powered  trans- 
mitters and  audio  equipment.  At  the  center  is  a hermetically  sealed  inductance  for  use 
under  poor  environmental  conditions . To  the  left  is  an  inexpensive  receiving-type  choke, 
with  a small  iron-core  r-f  choke  directly  in  front  of  it. 


CHAPTER  THREE 


Alternating  Current  Circuits 


The  previous  chapter  has  been  devoted  to 
a discussion  of  circuits  and  circuit  elements 
upon  which  is  impressed  a current  consisting 
of  a flow  of  electrons  in  one  direction.  This 
type  of  unidirectional  cuttent  flow  is  called 
direct  curtent,  abbreviated  d.  c.  Equally  as  im- 
portant in  radio  and  communications  work, 
and  power  practice,  is  a type  of  curtent  flow 
whose  direction  of  electron  flow  reverses 
periodically.  The  reversal  of  flow  may  take 
place  at  a low  rate,  in  the  case  of  power  sys- 
tems, or  it  may  take  place  millions  of  times 
per  second  in  the  case  of  communications 
frequencies.  This  type  of  current  flow  is 
called  alternating  current,  abbreviated  a.  c. 


3-1  Alternating  Current 


Frequency  of  on  An  alternating  current  is 
Alternating  Current  one  whose  amplitude  of 
current  flow  periodically 
rises  from  zero  to  a maximum  in  one  direction, 
decreases  to  zero,  changes  its  direction, 
rises  to  maximum  in  the  opposite  direction, 
and  decreases  to  zero  again.  This  complete 
process,  starting  from  zero,  passing  through 
two  maximums  in  opposite  directions,  and  re- 
turning to  zero  again,  is  called  a cycle.  The 
number  of  times  per  second  that  a current 
passes  through  the  complete  cycle  is  called 
the  frequency  of  the  current.  One  and  one 
quarter  cycles  of  an  alternating  curtent  wave 
are  illustrated  diagtammatically  in  figure  1. 


Frequency  Spectrum  At  present  the  usable  fre- 
quency range  for  alternat- 
ing electrical  currents  extends  over  the  enor- 
mous frequency  range  from  about  15  cycles  per 
second  to  perhaps  30,000,000,000  cycles  per 
second.  It  is  obviously  cumbersome  to  use  a 
frequency  designation  in  c.p.s.  for  enormously 
high  frequencies,  so  three  common  units  which 
are  multiples  of  one  cycle  per  second  have 
been  established. 


Figure  1 

ALTERNATING  CURRENT 
AND  DIRECT  CURRENT 

Graphical  comparison  bsiwaon  unidractionai 
(diracti  currant  and  altarnating  currant  as  plotted 
against  time. 


41 


42  Alternating  Current  Circuits 


THE  RADIO 


These  units  are: 

(1)  the  kilocycle  (abbr.,  kc.),  1000  c.p.s. 

(2)  the  Megacycle  (abbr.,  Me.),  1,000,000 
c.p.s.  or  1000  kc. 

(3)  the  kilo-Megacycle  (abbr.,  kMc.), 
1,000,000,000  c.p.s.  or  1000  Me. 

With  easily  handled  units  such  as  these  we 
can  classify  the  entire  usable  frequency  range 
into  frequency  bands. 

The  frequencies  falling  between  about  15 
and  20,000  c.p.s.  ate  called  audio  frequencies, 
abbreviated  a.j.,  since  these  frequencies  ate 
audible  to  the  human  ear  when  converted  from 
electrical  to  acoustical  signals  by  a loud- 
speaker or  headphone.  Frequencies  in  the 
vicinity  of  60  c.p.s.  also  ate  called  power  fre- 
quencies, since  they  ate  commonly  used  to 
distribute  electrical  power  to  the  consumer. 

The  frequencies  falling  between  10,000 
c.p.s.  (10  kc.)  and  30,000,000,000  c.p.s.  (30 
kMc.)  ate  commonly  called  radio  frequencies, 
abbreviated  r.f.,  since  they  are  commonly  used 
in  radio  communication  and  allied  arts.  The 
radio-frequency  spectrum  is  often  arbitrarily 
classified  into  seven  frequency  bands,  each 
one  of  which  is  ten  times  as  high  in  frequency 
as  the  one  just  below  it  in  the  spectrum  (ex- 
cept for  the  v-l-f  band  at  the  bottom  end  of 
the  spectrum).  The  present  spectrum,  with 
classifications,  is  given  below. 


Frequency  Classification  Abbrev. 

10  to  30  kc.  Very-low  frequencies  v.l.f 

30  to  300  kc.  Low  frequencies  l.f. 

300  to  3000  kc.  Medium  frequencies  m.f. 

3 to  30  Me.  High  frequencies  h.f. 


30  to  300  Me.  Very-high  frequencies  v.h.f. 
300  to  3000  Me.  Ultra-high  frequencies  u.h.f. 
3 to  30  kMc.  Super-high  frequencies  s.h.f. 
30  to  300  kMc.  Exttemely-high 

frequencies  e.h.f. 

Generation  of  Faraday  discovered  that  if 

Alternating  Current  a conductor  which  forms 
part  of  a closed  circuit  is 
moved  through  a magnetic  field  so  as  to  cut 
across  the  lines  of  force,  a current  will  flow 
in  the  conductor.  He  also  discovered  that,  if  a 
conductor  in  a second  closed  circuit  is  brought 
neat  the  first  conductor  and  the  current  in  the 
first  one  is  varied,  a current  will  flow  in  the 
second  conductor.  This  effect  is  known  as 
induction,  and  the  currents  so  generated  are 
induced  currents.  In  the  latter  case  it  is  the 
lines  of  force  which  are  moving  and  cutting 
the  second  conductor,  due  to  the  varying  cur- 
rent strength  in  the  first  conductor. 

A current  is  induced  in  a conductor  if  there 
is  a relative  motion  between  the  conductor 
and  a magnetic  field,  its  direction  of  flow  de- 
pending upon  the  direction  of  the  relative 


5em/>scHemof/c  represanfaf/on  of  fhe  simplest 
form  of  the  alternator. 


motion  between  the  conductor  and  the  field, 
and  its  strength  depends  upon  the  intensity  of 
the  field,  the  rate  of  cutting  lines  of  force,  and 
the  number  of  turns  in  the  conductor. 

Alternators  A machine  that  generates  an  alter- 
nating current  is  called  an  alter- 
nator or  a-c  generator.  Such  a machine  in  its 
basic  form  is  shown  in  figure  2.  It  consists  of 
two  permanent  magnets,  M,  the  opposite  poles 
of  which  face  each  other  and  are  machined  so 
that  they  have  a common  radius.  Between 
these  two  poles,  north  (N)  and  south  (S), 
a substantially  constant  magnetic  field  exists. 
If  a conductor  in  the  form  of  C is  suspended 
so  that  it  can  be  freely  rotated  between  the 
two  poles,  and  if  the  opposite  ends  of  con- 
ductor C are  brought  to  collector  rings,  there 
will  be  a flow  of  Cremating  current  when  con- 
ductor C is  rotated.  This  current  will  flow  out 
through  the  collector  rings  R and  brushes  B 
to  the  external  circuit,  X-Y. 

The  field  intensity  between  the  two  pole 
pieces  is  substantially  constant  over  the  entire 
area  of  the  pole  face  However,  when  the 
conductor  is  moving  parallel  to  the  lines  of 
force  at  the  top  or  bottom  of  the  pole  faces, 
no  lines  are  being  cut.  As  the  conductor  moves 
on  across  the  pole  face  it  cuts  more  and  more 
lines  of  force  for  each  unit  distance  of  travel, 
until  it  is  cutting  the  maximum  number  of 
lines  when  opposite  the  center  of  the  pole. 
Therefore,  zero  current  is  induced  in  the  con- 
ductor at  the  instant  it  is  midway  between 
the  two  poles,  and  maximum  current  is  in- 
duced when  it  is  opposite  the  center  of  the 
pole  face.  After  the  conductor  has  rotated 
through  180°  it  can  be  seen  that  its  position 
with  respect  to  the  pole  pieces  will  be  exactly 
opposite  to  that  when  it  started.  Hence,  the 
second  180°  of  rotation  will  produce  an  alter- 
nation of  current  in  the  opposite  direction  to 
that  of  the  first  alternation. 

The  current  does  not  increase  directly  as 
the  angle  of  rotation,  but  rather  as  the  sine 
of  the  angle;  hence,  such  a current  has  the 
mathematical  form  of  a sine  wave.  Although 


HANDBOOK 


The  Sine  Wave  43 


LINES  OF  FORCE 


Figure  3 

OUTPUT  OF  THE  ALTERNATOR 

Graph  showing  slna-wavm  output  currant  of  tha 
altarnator  of  flgura  2. 


most  electrical  machinery  does  not  produce  a 
strictly  pure  sine  curve,  the  departures  are 
usually  so  slight  that  the  assumption  can  be 
regarded  as  fact  for  most  practical  purposes. 
All  that  has  been  said  in  the  foregoing  para- 
graphs concerning  alternating  current  also  is 
applicable  to  alternating  voltage. 

The  rotating  arrow  to  the  left  in  figure  3 
represents  a conductor  rotating  in  a constant 
magnetic  field  of  uniform  density.  The  arrow 
also  can  be  taken  as  a vector  representing  the 
strength  of  the  magnetic  field.  This  means  that 
the  length  of  the  arrow  is  determined  by  the 
strength  of  the  field  (number  of  lines  of  force), 
which  is  constant.  Now  if  the  arrow  is  rotating 
at  a constant  rate  (that  is,  with  constant 
angular  velocity),  then  the  voltage  developed 
across  the  conductor  will  be  proportional  to 
the  rate  at  which  it  is  cutting  lines  of  force, 
which  rate  is  proportional  to  the  vertical 
distance  between  the  tip  of  the  arrow  and  the 
horizontal  base  line. 

If  EO  is  taken  as  unity  or  a voltage  of  1, 
then  the  voltage  (vertical  distance  from  tip  of 
arrow  to  the  horizontal  base  line)  at  point  C 
for  instance  may  be  determined  simply  by 
referring  to  a table  of  sines  and  looking  up  the 
sine  of  the  angle  which  the  arrow  makes  with 
the  horizontal. 

When  the  arrow  has  traveled  from  A to  point 
E,  it  has  traveled  90  degrees  or  one  quarter 
cycle.  The  other  three  quadrauits  are  not  shown 
because  their  complementary  or  mirror  relation- 
ship to  the  first  quadrant  is  obvious. 

It  is  important  to  note  that  time  units  are 
represented  by  degrees  or  quadrants.  The  fact 
that  AB,  BC,  CD,  and  DE  are  equal  chords 
(forming  equal  quadrants)  simply  means  that 
the  arrow  (conductor  or  vector)  is  traveling 
at  a constant  speed,  because  these  points  on 
the  radius  represent  the  passage  of  equal 
units  of  time. 

The  whole  picture  can  be  represented  in 
another  way,  and  its  derivation  from  the  fore- 
going is  shown  in  figure  3-  The  time  base  is 
represented  by  a straight  line  rather  than  by 


Figure  4 

THE  SINE  WAVE 

llluttrating  one  cycle  of  o s/ne  wove.  One 
complete  cycle  of  alternation  Is  broken  up 
Into  360  degrees.  Then  one-half  cycle  Is  180 
degrees,  one-quarter  cycle  Is  90  degrees,  and 
so  on  down  to  the  smallest  division  of  the 
wove.  A cosine  wove  has  a shape  Identical  to 
a sine  wove  but  Is  shifted  90  degrees  In  phase 
In  other  words  the  wave  begins  at  full  am- 
plitude, the  90-degree  point  comes  of  zero  am- 
plitude, the  180-degree  point  comes  at  full 
amplitude  In  the  opposite  direction  of  current 
flow,  etc. 


angular  rotation.  Points  A,  B,  C,  etc.,  rep- 
resent the  same  units  of  time  as  before.  When 
the  voltage  corresponding  to  each  point  is 
projected  to  the  corresponding  time  unit,  the 
familiar  sine  curve  is  the  result. 

The  frequency  of  the  generated  voltage  is 
proportional  to  the  speed  of  rotation  of  the 
alternator,  and  to  the  number  of  magnetic  poles 
in  the  field.  Alternators  may  be  built  to  produce 
radio  frequencies  up  to  30  kilocycles,  and 
some  such  machines  are  still  used  for  low 
frequency  communication  purposes.  By  means 
of  multiple  windings,  three-phase  output  may 
be  obtained  from  large  industrial  alternators. 

Radian  Notatian  From  figure  1 we  see  that  the 
value  of  an  a-c  wave  varies 
continuously.  It  is  often  of  importance  to  know 
the  amplitude  of  the  wave  in  terms  of  the 
total  amplitude  at  any  instant  or  at  any  time 
within  the  cycle.  To  be  able  to  establish  the 
instant  in  question  we  must  be  able  to  divide 
the  cycle  into  parts.  We  could  divide  the  cycle 
into  eighths,  hundredths,  or  any  other  ratio  that 
suited  our  fancy.  However,  it  is  much  more 
convenient  mathematically  to  divide  the  cycle 
either  into  electrical  degrees  (360°  represent 
one  cycle)  or  into  radians.  A radian  is  an  arc 
of  a circle  equal  to  the  radius  of  the  circle; 
hence  there  are  2n  radians  per  cycle  — or  per 
circle  (since  there  are  n diameters  per  circum- 
ference, there  are  2n-  radii). 

Both  radian  notation  and  electrical  degree 


44  Alternating  Current  Circuits 


THE  RADIO 


notation  are  used  in  discussions  of  alternating 
current  circuits.  However,  trigonometric  tables 
are  much  more  readily  available  in  terms  of 
degrees  than  radians,  so  the  following  simple 
conversions  are  useful. 

2n  radians  = 1 cycle  = 360° 
n radians  = \ cycle  = 180° 


77 

— radians  = ‘4  cycle  = 


90° 


77 

— radians  = \ cycle  = 


60° 


n 

— radians  = \ cycle  = 


45° 


1 

1 radian  = cycle  = 57.3° 

277 


When  the  conductor  in  the  simple  alter- 
nator of  figure  2 has  made  one  complete  revo- 
lution it  has  generated  one  cycle  and  has  ro- 
tated through  277  radians.  The  expression  27rf 
then  represents  the  number  of  radians  in  one 
cycle  multiplied  by  the  number  of  cycles  per 
second  (the  frequency)  of  the  alternating 
voltage  or  current.  The  expression  then  repre- 
sents the  number  of  radians  per  second  through 
which  the  conductor  has  rotated.  Hence  27Tf 
represents  the  angular  velocity  of  the  rotating 
conductor,  or  of  the  rotating  vector  which 
represents  any  alternating  current  or  voltage, 
expressed  in  radians  per  second. 

In  technical  literature  the  expression  277  f 
is  often  replaced  by  <u,  the  lower-case  Greek 
letter  omega.  Velocity  multiplied  by  time 
gives  the  distance  travelled,  so  27Tft  (or  &)t) 
represents  the  angular  distance  through  which 
the  rotating  conductor  or  the  rotating  vector 
has  travelled  since  the  reference  time  t = 0. 
In  the  case  of  a sine  wave  the  reference  time 
t = 0 represents  that  instant  when  the  voltage 
or  the  current,  whichever  is  under  discussion, 
also  is  equal  to  zero. 


WHERE: 

e (theta)  = PHASEANGLE  = 27rFT 
A = RADIANS  OR  90° 

B = 7r  RADIANS  OR  160* 

C = -^  RADIANS  OR  270» 

D = aTT  RADIANS  OR  360* 

1 RADIAN  = 57.324  DEGREES 


Figure  5 

ILLUSTRATING  RADIAN  NOTATION 

The  radian  is  a unit  of  phase  angle,  equal  to 
57.324  degrees.  It  is  commonly  used  in  mathe- 
matical relationships  involving  phase  angles 
since  such  re(of/ons/i/ps  ore  simplified  when 
radian  notation  is  used. 


where  e = the  instantaneous  voltage 

E = maximum  crest  value  of  voltage, 

f = frequency  in  cycles  per  second,  and 

t = period  of  time  which  has  elapsed 
since  t = 0 expressed  as  a fraction 
of  one  second. 

The  instantaneous  current  can  be  found  from 
the  same  expression  by  substituting  i for  e 
and  Imax  for  Emax. 

It  is  often  easier  to  visualize  the  process  of 
determining  the  instantaneous  amplitude  by 
ignoring  the  frequency  and  considering  only 
one  cycle  of  the  a-c  wave.  In  this  case,  for  a 
sine  wave,  the  expression  becomes; 

e = Emax  sin  0 

where  6 represents  the  angle  through  which 
the  vector  has  rotated  since  time  (and  ampli- 
tude) were  zero.  As  examples: 

when  6 = 30° 
sin  0 = 0.5 
so  e = 0.5  Emax 


Instantaneous  Value  The  instantaneous  volt- 
of  Valtage  or  age  or  current  is  propor- 

Current  tional  to  the  sine  of  the 

angle  through  which  the 
rotating  vector  has  travelled  since  reference 
time  t = 0.  Hence,  when  the  peak  value  of  the 
a-c  wave  amplitude  (either  voltage  or  cur- 
rent amplitude)  is  known,  and  the  angle  through 
which  the  rotating  vector  has  travelled  is 
established,  the  amplitude  of  the  wave  at 
this  instant  can  be  determined  through  use 
of  the  following  expression: 

e - Emax  sin  277ft, 


when  0 = 60° 
sin  0 = 0.866 
so  e = 0.866  Emax 


when  0 = S>0° 
sin  0 = 1.0 
so  e = Emax 


when  0=1  radian 

sin  0 = 0.8415 
so  e = 0.8415  Emax 


HANDBOOK 


A-C  Relationships  45 


Effective  Value  The  instantaneous  value 

of  an  of  an  alternating  current 

Alternating  Current  or  voltage  varies  continu- 
ously throughout  the  cycle. 
So  some  value  of  an  a-c  wave  must  be  chosen 
to  establish  a relationship  between  the  effec- 
tiveness of  an  a-c  and  a d-c  voltage  or  cur- 
rent; The  heating  value  of  an  alternating 
current  has  been  chosen  to  establish  the  refer- 
ence between  the  effective  values  of  a.c.  and 
d.  c.  Thus  an  alternating  current  will  have  an 
effective  value  of  1 ampere  when  it  produces 
the  same  heat  in  a resistor  as  does  I ampere 
of  direct  current. 

The  effective  value  is  derived  by  taking  the 
instantaneous  values  of  current  over  a cycle  of 
alternating  current,  squaring  these  values, 
taking  an  average  of  the  squares,  and  then 
taking  the  square  root  of  the  average.  By  this 
procedure,  the  effective  value  becomes  known 
as  the  root  mean  square  or  r.m.s.  value.  This 
is  the  value  that  is  read  on  a-c  voltmeters  and 
a-c  ammeters.  The  r.m.s.  value  is  70.7  (for 
sine  waves  only)  per  cent  of  the  peak  or  maxi- 
mum instantaneous  value  and  is  expressed  as 
follows: 

E eff.  or  Ec. m.  s.  = 0.707  x Emax  or 
leff.  or  If.m.s.  = 0.707  X Imax. 

The  following  relations  are  extremely  useful 
in  radio  and  power  work: 

Ef.m.s.  = 0.707  X Emax,  and 

Emax  = 1.414  X Er.m.s. 

Rectified  Alternating  If  an  alternating  current 
Current  or  Pulsat-  is  passed  through  a reed- 
ing Direct  Current  fier,  it  emerges  in  the 

form  of  a current  of 
varying  amplitude  which  flows  in  one  direc- 
tion only.  Such  a current  is  known  as  rectified 
a.c.  or  pulsating  d.c.  A typical  wave  form  of  a 
pulsating  direct  current  as  would  be  obtained 
from  the  output  of  a full-wave  rectifier  is 
shown  in  figure  6. 

Measuring  instruments  designed  for  d-c 
operation  will  not  read  the  peak  for  instantan- 
eous maximum  value  of  the  pulsating  d-c  out- 
put from  the  rectifier;  they  will  read  only  the 
average  value.  This  can  be  explained  by  as- 
suming that  it  could  be  possible  to  cut  off 
some  of  the  peaks  of  the  waves,  using  the  cut- 
off portions  to  fill  in  the  spaces  that  are  open, 
thereby  obtaining  an  average  d-c  value.  A 
milliammeter  and  voltmeter  connected  to  the 
adjoining  circuit,  or  across  the  output  of  the 
rectifier,  will  read  this  average  value.  It  is  re- 
lated to  peak  value  by  the  following  expres- 
sion: 

Eavg  = 0.636  X Emax 


Figure  6 

FULL-WAVE  RECTIFIED 
SINE  WAVE 

Woveform  obtained  at  the  output  of  a fullweve 
rectifier  being  fed  with  a sine  wave  and  baying 
100  per  cent  rectification  efficiency.  Each 
pulse  has  the  same  shape  as  one^half  cycle  of 
a sine  wove.  This  type  of  current  is  known  as 
pulsating  direct  current. 


It  is  thus  seen  that  the  average  value  is  63.6 
per  cent  of  the  peak  value. 

Relationship  Between  To  summarize  the  three 
Peak,  R.M.S.  or  most  significant  values 

Effective,  ond  of  an  a-c  sine  wave:  the 

Average  Values  peak  value  is  equal  to 

1.41  times  the  r.m.s.  or 
effective,  and  the  r.m.s.  value  is  equal  to 
0.707  times  the  peak  value;  the  average  value 
of  a full-wave  rectified  a-c  wave  is  0.636 
times  the  peak  value,  and  the  average  value 
of  a rectified  wave  is  equal  to  0.9  times  the 
r.m.s.  value. 

R.M.S.  = 0.707  X Peak 

Average  = 0.636  x Peak 


Average  = 0.9  x R.M.S. 
R.M.S.  =1.11  X Average 


Peak  = 1.414  x R.M.S. 

Peak  =1.57  x Average 

Applying  Ohm’s  Law  Ohm’s  law  applies 
to  Alternating  Current  equally  to  direct  or  al- 
ternating current,  pro- 
vided the  circuits  under  consideration  are 
purely  resistive,  that  is,  circuits  which  have 
neither  inductance  (coils)  nor  capacitance 
(capacitors).  Problems  which  involve  tube 
filaments,  drop  resistors,  electric  lamps, 
heaters  or  similar  resistive  devices  can  be 
solved  from  Ohm’s  law,  regardless  of  whether 
the  current  is  direct  or  iternating.  When  a 
capacitor  or  coil  is  made  a part  of  the  circuit, 
a property  common  to  either,  called  reactance, 
must  be  taken  into  consideration.  Ohm’s  law 
still  applies  to  a-c  circuits  containing  react- 
ance,  but  additional  considerations  are  in- 
volved; these  will  be  discussed  in  a later 
paragr^h. 


46  Alternating  Current  Circuits 


THE  RADIO 


Figure  7 

LAGGING  PHASE  ANGLE 

Showing  the  manner  In  which  the  current  lags 
the  voltage  In  an  a-c  circuit  containing  pure 
Inductance  only-  The  lag  Is  equal  to  one-quarter 
cycle  or  90  degrees. 


Figure  8 

LEADING  PHASE  ANGLE 

Showing  the  monner  In  which  the  current  leads 
the  voltage  In  an  a-e  circuit  containing  pure 
capacitance  only.  The  lead  Is  equal  to  one- 
quarter  cycle  or  90  degrees. 


Inductive  As  was  stated  in  Chapter  Two, 

Reactance  when  a changing  current  flows 
through  an  inductor  a back-  or 
counter-electromotive  force  is  developed; 
opposing  any  change  in  the  initial  current. 
This  property  of  an  inductor  causes  it  to  offer 
opposition  or  impedance  to  a change  in  cur- 
rent. The  measure  of  impedance  offered  by  an 
inductor  to  an  alternating  current  of  a given 
frequency  is  known  as  its  inductive  reactance. 
This  is  expressed  as  Xj,. 

Xl  = 2»7fL, 

where  XL  = inductive  reactance  expressed  in 
ohms. 

77  = 3-1416  (2rr  = 6.283), 

f = frequency  in  cycles, 

L = inductance  in  henrys. 

Inductive  Reactance  It  is  very  often  neces- 
at  Radio  Frequencies  sary  to  compute  induc- 
tive reactance  at  radio 
frequencies.  The  same  formula  may  be  used, 
but  to  make  it  less  cumbersome  the  inductance 
is  expressed  in  millihenrys  and  the  frequency 
in  kilocycles.  For  higher  frequencies  and 
smaller  values  of  inductance,  frequency  is 
expressed  in  megacycles  and  inductance  in 
microhenrys.  The  basic  equation  need  not  be 
changed,  since  the  multiplying  factors  for 
inductance  and  frequency  appear  in  numerator 
and  denominator,  and  hence  are  cancelled  out. 
However,  it  is  not  possible  in  the  same  equa- 
tion to  express  Lin  millihenrys  and  f in  cycles 
without  conversion  factors. 

Capacitive  It  has  been  explained  that  induc- 
Reactance  tive  reactance  is  the  measure  of 
the  ability  of  an  inductor  to  offer 
impedance  to  the  flow  of  an  alternating  current. 


Capacitors  have  a similar  property  although 
in  this  case  the  opposition  is  to  any  change  in 
the  voltage  across  the  capacitor.  This  property 
is  called  capacitive  reactance  and  is  ex- 
pressed as  follows; 


where  Xc  = capacitive  reactance  in  ohms, 

77=  3.1416 

f = frequency  in  cycles, 

C = capacitance  in  farads. 

Capacitive  Re-  Here  again,  as  in  the  case 

octance  at  of  inductive  reactance,  the 

Radio  Frequencies  units  of  capacitance  and 
frequency  can  be  converted 
into  smaller  units  for  practical  problems  en- 
countered in  radio  work.  The  equation  may 
be  written: 

1,000,000 

Xc= ^ , 

277fC 

where  f = frequency  in  megacycles, 

C = capacitance  in  micro-microfarads. 

In  the  audio  range  it  is  often  convenient  to 
express  frequency  (f)  in  cycles  and  c^ac- 
itance  (C)  in  microfarads,  in  which  event  the 
same  formula  applies. 

Phase  When  an  alternating  current  flows 
through  a purely  resistive  circuit,  it 
will  be  found  that  the  current  will  go  through 
maximum  and  minimum  in  perfect  step  with 
the  voltage.  In  this  case  the  current  is  said  to 
be  in  step  or  in  phase  with  the  voltage.  For 
this  reason.  Ohm’s  law  will  apply  equally  well 
for  a.c.  or  d.c.  where  pure  resistances  ate  con- 
cerned, provided  that  the  same  values  of  the 


HANDBOOK 


Reactance  47 


wave  (either  peak  or  r.m.s.)  for  both  voltage 
and  current  are  used  in  the  calculations. 

However,  in  calculations  involving  alternat- 
ing currents  the  voltage  and  current  are  not 
necessarily  in  phase.  The  current  through  the 
circuit  may  lag  behind  the  voltage,  in  which 
case  the  current  is  said  to  have  lagging  phase. 
Lagging  phase  is  caused  by  inductive  react- 
ance. If  the  current  reaches  its  maximum  value 
ahead  of  the  voltage  (figure  8)  the  current  is 
said  to  have  a leading  phase.  A leading  phase 
angle  is  caused  by  capacitive  reactance. 

In  an  electrical  circuit  containing  reactance 
only,  the  current  will  either  lead  or  lag  the 
voltage  by  90°.  If  the  circuit  contains  induc- 
tive reactance  only,  the  current  will  lag  the 
voltage  by  90°.  If  only  capacitive  reactance  is 
in  the  circuit,  the  current  will  lead  the  voltage 
by  90°. 

Reactances  Inductive  and  capacitive  re- 
in Combination  actance  have  exactly  opposite 
effects  on  the  phase  relation 
between  current  and  voltage  in  a circuit. 
Hence  when  they  are  used  in  combination 
their  effects  tend  to  neutralize.  The  combined 
effect  of  a capacitive  and  an  inductive  react- 
ance is  often  called  the  net  reactance  of  a 
circuit.  The  net  reactance  (X)  is  found  by  sub- 
tracting the  capacitive  reactance  from  the  in- 
ductive reactance,  X = Xl  - Xc. 

The  tesult  of  such  a combination  of  pure 
reactances  may  be  either  positive,  in  which 
case  the  positive  reactance  is  greater  so  that 
the  net  reactance  is  inductive,  or  it  may  be 
negative  in  which  case  the  capacitive  react- 
ance is  greater  so  that  the  net  reactance  is 
capacitive.  The  net  reactance  may  also  be 
zero  in  which  case  the  circuit  is  said  to  be 
resonant.  The  condition  of  resonance  will  be 
discussed  in  a later  section.  Note  that  induc- 
tive reactance  is  always  taken  as  being  posi- 
tive while  capacitive  reactance  is  always 
taken  as  being  negative. 

Impedance;  Circuits  Pure  reactances  intro- 

Containing  Reactance  duce  a phase  angle  of 
and  Resistance  90°  between  voltage  and 

current;  pure  resistance 
introduces  no  phase  shift  between  voltage  and 
current.  Hence  we  cannot  add  a reactance  and 
a resistance  directly.  When  a reactance  and  a 
resistance  are  used  in  combination  the  re- 
sulting phase  angle  of  current  flow  with  re- 
spect to  the  impressed  voltage  lies  somewhere 
between  plus  or  minus  90°  and  0°  depending 
upon  the  relative  magnitudes  of  the  reactance 
and  the  resistance. 

The  term  impedance  is  a general  term  which 
can  be  applied  to  any  electrical  entity  which 
impedes  the  flow  of  current.  Hence  the  term 
may  be  used  to  designate  a resistance,  a pure 


Y-axis 


Operation  on  tho  vector  (+A)  by  tbo  quantity  f— 1) 
cousos  vector  to  rotate  through  160  dogreos. 


reactance,  or  a complex  combination  of  both 
reactance  and  resistance.  The  designation  for 
impedance  is  Z.  An  impedance  must  be  de- 
fined in  such  a manner  that  both  its  magnitude 
and  its  phase  angle  ate  established.  The 
designation  may  be  accomplished  in  either  of 
two  ways  — one  of  which  is  convertible  into 
the  other  by  simple  mathematical  operations. 

The  "J”  Operator  The  first  method  of  des- 
ignating an  impedance  is 
actually  to  specify  both  the  resistive  and  the 
reactive  component  in  the  form  R + jX.  In  this 
form  R represents  the  resistive  component  in 
ohms  and  X represents  the  reactive  component. 
The  "j”  merely  means  that  the  X component 
is  reactive  and  thus  cannot  be  added  directly 
to  the  R component.  Plus  jX  means  that  the 
reactance  is  positive  or  inductive,  while  if 
minus  jX  were  given  it  would  mean  that  the 
reactive  component  was  negative  or  capacitive. 

In  figure  9 we  have  a vector  (+A)  lying  along 
the  positive  X-axis  of  the  usual  X-Y  coordi- 
nate system.  If  this  vector  is  multiplied  by  the 
quantity  (-1),  it  becomes  (-A)  and  its  position 
now  lies  along  the  X-axis  in  the  negative 
direction.  The  operator  (-1)  has  caused  the 
vector  to  rotate  through  an  angle  of  180  de- 
grees. Since  (-1)  is  equal  to  (y^-1  x \f^),  the 
same  result  may  be  obtained  by  operating  on 
the  vector  with  the  operator  X \/-l). 

However  if  the  vector  is  operated  on  but  once 
by  the  operator  ( \/~l),  it  is  caused  to  rotate 
only  90  degrees  (figure  10).  Thus  the  operator 
( V-Drotates  a vector  by  90  degrees.  For  con- 
venience, this  operator  is  called  the  j operator. 
In  like  fashion,  the  operator  (-j)  rotates  the 
vector  of  figure  9 through  an  angle  of  270 
degrees,  so  that  the  resulting  vector  (-jA) 
falls  on  the  (-Y)axis  of  the  coordinate  system. 


48  Alternating  Current  Circuits 


THE  RADIO 


Y-axis 


Figure  10 

Operation  on  the  vector  (+A}  by  the  quantity  (I) 
causes  vecfor  to  rotate  through  90  degrees. 

Polar  Notation  The  second  method  of  repre- 
senting an  impedance  is  to 
specify  its  absolute  magnitude  and  the  phase 
angle  of  current  with  respect  to  voltage,  in 
the  form  "L  LQ.  Figure  11  shows  graphically 
the  relationship  between  the  two  common  ways 
of  representing  an  impedance. 

The  construction  of  figure  11  is  called  an 
impedance  diagram.  Through  the  use  of  such 
a i^agtam  we  can  add  graphically  a resistance 
and  a reactance  to  obtain  a value  for  the  re- 
sulting impedance  in  the  scalar  form.  With 
zero  at  the  origin,  resistances  are  plotted  to 
the  right,  positive  values  of  reactance  (induc- 
tive) in  the  upward  direction,  and  negative 
values  of  reactance  (capacitive)  in  the  down- 
ward direction. 

Note  that  the  resistance  and  reactance  are 
drawn  as  the  two  sides  of  a right  triangle, 
with  the  hypotenuse  representing  the  resulting 
impedance.  Hence  it  is  possible  to  deter- 
mine mathematically  the  value  of  a resultant 
impedance  through  the  familiar  right-triangle 
relationship  — the  square  of  the  hypotenuse  is 
equal  to  the  sum  of  the  squares  of  the  other 
two  sides: 

Z'  = R"  + X" 
or  jZj  = + X' 

Note  also  that  the  angle  6 included  between 
R and  Z can  be  determined  from  any  of  the 
following  trigonometric  relationships; 


X 


R 

cos  0 = -— - 

izl 

X 

tan  0 = — 

R 

One  common  problem  is  that  of  determining 
the  scalar  magnitude  of  the  impedance,  |Zj, 


Figure  1 1 

THE  IMPEDANCE  TRIANGLE 

Showing  the  graphical  construction  of  a triangle 
for  obtaining  the  net  f$ca/ar)  impeciance  re> 
suiting  from  the  connection  of  a resistance  and 
a reactance  in  series.  Shown  also  alongside  Is 
the  o/ternofrve  inafhemaf/ca/  procedure  for 
obtaining  the  values  associated  with  the  triangle. 


and  the  phase  angle  0,  when  resistance  and 
reactance  ate  known;  hence,  of  converting 
from  the  Z = R -1  jX  to  the  jZ]  Z.  0 form.  In  this 
case  we  use  two  of  the  expressions  just  given: 

Izl  = V R'  + X' 

X X 

tan  0 = — , ( or  0 = tan"’  — ) 

R R 

The  inverse  problem,  that  of  converting 
from  the  jZ]  Z.  0 to  the  R + jX  form  is  done 
with  the  following  relationships,  both  of  which 
are  obtainable  by  simple  division  from  the 
trigonometric  expressions  just  given  for  de- 
termining the  angle  0: 

R = jZj  cos  0 
jX  = jZj  j sin  0 

By  simple  addition  these  two  expressions  may 
be  combined  to  give  the  relationship  between 
the  two  most  common  methods  of  indicating 
an  impedance; 

R + jX  = \7.\  (cos  0 -I-  j sin  0) 

In  the  case  of  impedance,  resistance,  or  re- 
actance, the  unit  of  measurement  is  the  ohm; 
hence,  the  ohm  may  be  thought  of  as  a unit  of 
opposition  to  current  flow,  without  reference 
to  the  relative  phase  angle  between  the  ap- 
plied voltage  and  the  current  which  flows. 

Further,  since  both  capacitive  and  inductive 
reactance  ate  functions  of  frequency,  imped- 
ance will  vary  with  frequency.  Figure  12 
shows  the  manner  in  which  \X\  will  vary  with 
frequency  in  an  RL  series  circuit  and  in  an 
RC  series  circuit. 

Series  RLC  Circuits  In  a series  circuit  contain- 
ing R,  L,  and  C,  the  im- 


HANDBOOK 


Impedance  49 


pedance  is  determined  as  discussed  before  ex“ 
cept  that  the  reactive  component  in  the  ex* 
pressions  becomes:  (The  net  reactance  — the 
difference  between  Xl  and  Xc*)  Hence  (Xl  — 
Xc)  may  be  substituted  for  X in  the  equations. 
Thus; 

|zl  = VR'+(Xl  - Xc)" 


6 = tan~‘ 


(Xl  - Xc) 
R 


A series  RLC  circuit  thus  may  present  an 
impedance  which  is  capacitively  reactive  if 
the  net  reactance  is  capacitive,  inductively 
reactive  if  the  net  reactance  is  inductive,  or 
resistive  if  the  capacitive  and  inductive  re- 
actances are  equal. 


Addition  of  The  addition  of  complex 

Complex  Quontities  quantities  (for  example, 

impedances  in  series)  is 
quite  simple  if  the  quantities  are  in  the  rect- 
angular form.  If  they  are  in  the  polar  form 
they  only  can  be  added  graphically,  unless 
they  are  converted  to  the  rectangular  form  by 
the  relationships  previously  given.  As  an  ex- 
ample of  the  addition  of  complex  quantities 
in  the  rectangular  form,  the  equation  for  the 
addition  of  impedances  is: 

(R.  + jX.)  + (Rj  -i-  jXj)=  (R,  4-  R,)  + KX,  + X,) 

For  example  if  we  wish  to  add  the  imped- 
ances (10+j50)  and  (20  - j30)  we  obtain: 

(10  + jSO)  (20  - j30) 

= (10  + 20)  + j(50  +(-30)) 

= 30  + j(50-30) 

= 30  + i20 


Multiplication  and  It  is  often  necessary  in 
Division  of  solving  certain  types  of 

Complex  Quantities  circuits  to  multiply  or  di- 
vide two  complex  quan- 
tities. It  is  a much  simplier  mathematical 
operation  to  multiply  or  divide  complex  quan- 
tities if  they  are  expressed  in  the  polar  form. 
Hence  if  they  are  given  in  the  rectangular 
form  they  should  be  converted  to  the  polar 
form  before  multiplication  or  division  is  begun. 
Then  the  multiplication  is  accomplished  by 
multiplying  the  |2|  terms  together  and  adding 
algebraically  the  L d terms,  as: 

(lz,j  Le,)(\z,\  Le,)  = |z.|  |z,|  (Ld,  + Ld,) 

For  example,  suppose  that  the  two  impedances 
|20|  A 43°  and  |32l  Z-23°  are  to  be  multi- 
plied. Then: 

( |20|  Z43°)  (l32l  Z-23°)  = |20-32| 

( Z43°  + 21-23°) 

= 640  L 20° 


Figure  12 

IMPEDANCE  AGAINST  FREQUENCY 
FOR  R.L  AND  R-C  CIRCUITS 

T/ie  impedance  of  on  R-C  circuit  approaches 
infinity  as  the  frequency  approaches  zero  (d.c.), 
while  the  impedance  of  a series  R-L  circuit 
approaches  Infinity  as  the  frequency  approaches 
infinity.  The  impedance  of  an  R-C  circuit  ap- 
proaches the  impedance  of  the  series  resistor 
as  the  frequency  opprooches  infinity,  while  the 
Impedance  of  a series  R-L  circuit  approaches 
the  Impedance  of  the  resistor  as  the  frequency 
approaches  zero. 


Division  is  accomplished  by  dividing  the 
denominator  into  the  numerator,  and  su6- 
true  ting  the  angle  of  the  denominator  from 
that  of  the  numerator,  as: 


\z,\Le,  |z/  ^ 


For  example,  suppose  that  an  impedance  of 
|50|  A 67°  is  to  be  divided  by  an  impedance 
of  [10 1 2145°.  Then: 


|50|  L67° 
|10|  Z45° 


|10| 


(Z  67° -A  45°)  = 15  i(Z22°) 


Ohm’s  Law  for  The  simple  form  of  Ohm’s 

Complex  Quantities  Law  used  for  d-c  circuits 
may  be  stated  in  a mote 
general  form  for  application  to  a-c  circuits 
involving  either  complex  quantities  or  simple 
resistive  elements.  The  form  is: 

E 

I = 

Z 

in  which,  in  the  general  case,  /,  E,  and  Z are 
complex  (vector)  quantities.  In  the  simple 
case  where  the  impedance  is  a pure  resistance 
with  an  a-c  voltage  applied,  the  equation 
simplifies  to  the  familiar  I = E/R.  In  any  case 

the  applied  voltage  may  be  expressed  either 
as  peak,  r.m.s.,  or  average;  the  resulting 


50  Alternating  Current  Circuits 


THE  RADIO 


o 


100  VOLTS 


O 


Figure  13 

SERIES  R-L-C  CIRCUIT 


current  always  will  be  in  the  same  type  of 
units  as  used  to  define  the  voltage. 

In  the  more  general  case  vector  algebra 
must  be  used  to  solve  the  equation.  And, 
since  either  division  or  multiplication  is  in- 
volved, the  complex  quantities  should  be  ex- 
pressed in  the  polar  form.  As  an  example, 
take  the  case  of  the  series  circuit  shown  in 
figure  13  with  100  volts  applied.  The  imped- 
ance of  the  series  circuit  can  best  be  obtained 
first  in  the  rectangular  form,  as; 

200  + j ( 100  - 300)  = 200  - j 200 

Now,  to  obtain  the  current  we  must  convert 
this  impedance  to  the  polar  form. 

iz]  = V 200^  +(-200)“ 

= V 40,000  + 40,000 
= v' 80,000 
= 282  0 


X -200 

6 = tan  ‘ — = tan"' = tan”  — 1 

R 200 


= -45° 

Therefore  Z = 282  ^-45° 

Note  that  in  a series  circuit  the  resulting  im- 
pedance takes  the  sign  of  the  largest  react- 
ance in  the  series  combination. 

Where  a slide-rule  is  being  used  to  make 
the  computations,  the  impedance  may  be  found 
without  any  addition  or  subtraction  operations 
by  finding  the  angle  6 first,  and  then  using 
the  trigonometric  equation  below  for  obtain- 
ing the  impedance.  Thus: 

X -200 

6 - tan"'  — = tan  ' = tan  - 1 

R 200 

= -45° 

Then  \Z\  = r co®  "45°  = 0.707 

cos  u 

200 

IZj  = =282n 

0.707 


Since  the  applied  voltage  will  be  the  reference 
for  the  currents  and  voltages  within  the  cir- 
cuit, we  may  define  it  as  having  a zero  phase 
angle;  E = 100  L0°  Then: 


I = 


100  Z.0  ° 

282  Z.-45° 


= 0.354  A0°-(-45°) 


= 0.35  4 2.45°  amperes. 


This  same  current  must  flow  through  all  three 
elements  of  the  circuit,  since  they  ate  in 
series  and  the  current  through  one  must  al- 
ready have  passed  through  the  other  two. 
Hence  the  voltage  drop  across  the  resistor 
(whose  phase  angle  of  course  is  0°)  is: 

E = 1 R 

E =(0.354  LA5°)  (200  Z0°) 

= 70.8  L 45°  volts 

The  voltage  drop  across  the  inductive  react- 
ance is: 


E = 1 Xl 

E = (0.354  /L45°)  (100  L90°) 

= 35.4  2.135°  volts 

Similarly,  the  voltage  drop  across  the  capac- 
itive reactance  is; 

E = /Xc 

E = (0.354  /-45°)  (300  /-90°) 

= 106.2  2-45° 

Note  that  the  voltage  drop  across  the  capac- 
itive reactance  is  greater  than  the  supply 
voltage.  This  condition  often  occurs  in  a 
series  RLC  circuit,  and  is  explained  by  the 
fact  that  the  drop  across  the  capacitive  react- 
ance is  cancelled  to  a lesser  or  greater  ex- 
tent by  the  drop  across  the  inductive  reactance. 

It  is  often  desirable  in  a problem  such  as 
the  above  to  check  the  validity  of  the  answer 
by  adding  vectorially  the  voltage  drops  across 
the  components  of  the  series  circuit  to  make 
sure  that  they  add  up  to  the  supply  voltage  — 
or  to  use  the  terminology  of  Kirchhoff’s  Second 
Law,  to  make  sure  that  the  voltage  drops 
across  all  elements  of  the  circuit,  including 
the  source  taken  as  negative,  is  equal  to  zero. 

In  the  general  case  of  the  addition  of  a 
number  of  voltage  vectors  in  series  it  is  best 
to  resolve  the  voltages  into  their  in-phase 
and  out-of-phase  components  with  respect  to 
the  supply  voltage.  Then  these  components 
may  be  added  directly.  Hence: 

Er  = 70.8  2 45° 

= 70.8  ( cos  45°  + j sin  45°) 

= 70.8  (0.707  + j0.707) 

= 50  + j50 


HANDBOOK 


Vector  Algebra  51 


90“ 


Figure  14 

Graphical  construcflan  of  tho  vo/foge  dropa 
assoclatod  with  iho  sorlos  R^L‘C  circuit  of 

figure  13. 


El  = 35.4  /.  135° 

= 35.4  ( cos  135°  + j sin  135°) 
= 35.4  (-0.707  + j0.707) 

= -25  +i25 


Ec  = 106.2  /45° 

= 106.2  ( cos-45°  + j sin-45°) 
= 106.2  (0.7O7  -j0.707) 

= 75  -j75 


£r  + El  + = (50  + j50)  + (-25  + j25) 

+ (75-j75) 

= (50  - 25  + 75)  + ) (50  + 25  - 75) 

= 100  + jO 

= 100  /0°,  which  is  equnl  to  the 
supply  voltage. 

Checking  by  It  is  frequently  desirable 

Construction  on  the  to  check  computations  in- 
Complex  Plane  volving  complex  quantities 

by  constructing  vectors 
representing  the  quantities  on  the  complex 
plane.  Figure  14  shows  such  a construction 
for  the  quantities  of  the  problem  just  com- 
pleted. Note  that  the  answer  to  the  problem 
may  be  checked  by  constructing  a ptuallel- 
ogram  with  the  voltage  drop  across  the  re- 
sistor as  one  side  and  the  net  voltage  drop 
across  the  capacitor  plus  the  inductor  (these 
may  be  added  algrebraically  as  they  are  180° 
out  of  phase)  as  the  adjacent  side.  The  vector 
sum  of  these  two  voltages,  i^yhich  is  repre- 
sented by  the  diagonal  of  the  parallelogram, 
is  equal  to  the  supply  voltage  of  100  volts  at 
zero  phase  angle. 


Resistance  and  Re-  In  a series  circuit,  such 
Qctoncg  in  Potollel  as  just  tliscussed,  the  cur- 
rent through  all  the  ele- 


PARALLEL CIRCUIT 


EQUIVALENT  SERIES  CIRCUIT 


Figure  15 

THE  EQUIVALENT  SERIES  CIRCUIT 

Showing  a parallel  R-C  circuit  and  the  egu/v* 
alent  seriee  R-C  circuit  which  represents  the 
same  net  Impedance  as  the  parallel  circuit. 


ments  which  go  to  make  up  the  series  circuit 
is  the  same.  But  the  voltage  drops  across 
each  of  the  components  ate,  in  general,  dif- 
ferent from  one  another.  Conversely,  in  a 
parallel  RLC  or  RX  circuit  the  voltage  is, 
obviously,  the  same  across  each  of  the  ele- 
ments. But  the  currents  through  each  of  the 
elements  ate  usually  different. 

There  are  many  ways  of  solving  a problem 
involving  paralleled  resistance  and  reactance; 
several  of  these  ways  will  be  described.  In 
general,  it  may  be  said  that  the  impedance  of 
a number  of  elements  in  parallel  is  solved 
using  the  same  relations  as  are  used  for 
solving  resistors  in  parallel,  except  that  com- 
plex quantities  are  employed.  The  basic  re- 
lation is: 


or  when  only  two  impedances  are  involved: 


Z 


lot- 


ZtZ, 
Zi  + Zj 


As  an  example,  using  the  two-impedance 
relation,  take  the  simple  case,  illustrated  in 
figure  15,  of  a resistance  of  6 ohms  in  paral- 
lel with  a capacitive  reactance  of  4 ohms.  To 
simplify  the  first  step  in  the  computation  it  is 
best  to  put  the  impedances  in  the  polar  form 
for  the  numerator,  since  multiplication  is  in- 
volved, and  in  the  rectangular  form  for  the 
addition  in  the  denominator. 


(6  /0°)  (4  1-90°) 


24  /-90° 

6-j4 

Then  the  denominator  is  changed  to  the  polar 
form  for  the  division  operation: 


6 = tan“ 


■ 0.667  = 


33.7° 


52  Alternating  Current  Circuits 


THE  RADIO 


z = - 


cos  - 33.7°  0.832 

6-j4  = 7.2lZ-33.r 


= 7.21  ohms 


Then: 


24 

7.21  A- 33.7° 


3.33  A-56.3° 


= 3-33  ( cos  - 56.3°  + j sin  -56.3°) 
= 3.33  [0.5548  + j (-0.832)1 
- 1.85  - j 2.77 


E2  = Ei 


R2 

R1  + R2 


@ 


E2=Ei 

E2  = Ei 


XCi  +XC2 
C2 


E2  = Ei 


L2 


© 


Figure  16 

SIMPLE  A-C  VOLTAGE  DIVIDERS 


Equivalent  Series  Through  the  series  of  op- 
Circuit  erations  in  the  previous 

paragraph  we  have  convert- 
ed a circuit  composed  of  two  impedances  in 
parallel  into  an  equivalent  series  circuit  com- 
posed of  impedances  in  series.  An  equivalent 
series  circuit  is  one  which,  as  far  as  the  ter- 
minals are  concerned,  acts  identically  to  the 
original  parallel  circuit;  the  current  through 
the  circuit  and  the  power  dissipation  of  the 
resistive  elements  are  the  same  for  a given 
voltage  at  the  specified  frequency. 

We  can  check  the  equivalent  series  circuit 
of  figure  15  with  respect  to  the  original  cir- 
cuit by  assuming  that  one  volt  a.c.  (at  the 
frequency  where  the  capacitive  reactance  in 
the  parallel  circuit  is  4 ohms)  is  applied  to 
the  terminals  of  both. 

In  the  parallel  circuit  the  current  through 
the  resistor  will  be  % ampere  (0. 166a.)  while 
the  current  through  the  capacitor  will  be  j \ 
ampere  (+  j 0.25  a.).  The  total  cutrent  will  be 
the  sum  of  these  two  currents,  or  0. 166  + 
j 0.25  a.  Adding  these  vectotially  we  obtain: 


in  = VO.166"  + 0.25"  = \/0-09  =0.3  a. 

The  dissipation  in  the  resistor  will  be  lV6  = 
0. 166  watts. 

In  the  case  of  the  equivalent  series  circuit 
the  current  will  be: 


111 


E 1 

|Zl  “ 3.33 


0.3  a. 


And  the  dissipation  in  the  resistor  will  be: 


W = I"R  = 0.3"  X 1.85 
= 0.9  X 1.85 
= 0. 166  watts 

So  we  see  that  the  equivalent  series  circuit 
checks  exactly  with  the  original  parallel  cir- 
cuit. 


Parallel  RLC  In  solving  a more  complicated 
Circuits  circuit  made  up  of  more  than 

two  impedances  in  parallel  we 


may  elect  to  use  either  of  two  methods  of 
solution.  These  methods  ate  called  the  admit- 
tance method  and  the  assumed- voltage  method. 
However,  the  two  methods  are  equivalent 
since  both  use  the  sum-of-reciprocals  equation: 


In  the  admittance  method  we  use  the  relation 
Y = 1/Z,  where  Y = G + jB;  Y is  called  the 
admittance,  defined  above,  G is  the  conduct- 
ance or  R/Z"  and  B is  the  susceptance  or 

-X/Z".ThenY,oi  = l/Z,o,  =Y,  +Y,-t-Y, 

In  the  assumed-voltage  method  we  multiply 
both  sides  of  the  equation  above  by  E,  the 
assumed  voltage,  and  add  the  currents,  as: 


= Iz.  + Iz.  +I2 


Then  the  impedance  of  the  parallel  com- 
bination may  be  determined  from  the  relation: 

^tot  = E/ 12  tot 


AC  Voltage  Voltage  dividers  for  use  with 
Dividers  alternating  current  ate  quite  simi- 
lar to  d-c  voltage  dividers.  How- 
ever, since  capacitors  and  inductors  oppose 
the  flow  of  a-c  cutrent  as  well  as  resistors, 
voltage  dividers  for  alternating  voltages  may 
take  any  of  the  configurations  shown  in  fig- 
ure 16. 

Since  the  impedances  within  each  divider 
are  of  the  same  type,  the  output  voltage  is  in 
phase  with  the  input  voltage.  By  using  com- 
binations of  different  types  of  impedances,  the 
phase  angle  of  the  output  may  be  shifted  in 
relation  to  the  input  phase  angle  at  the  same 
time  the  amplitude  is  reduced.  Several  di- 
viders of  this  type  ate  shown  in  figure  17. 
Note  that  the  ratio  of  output  voltage  to  input 
voltage  is  equal  to  the  ratio  of  the  output 
impedance  to  the  total  divider  impedance. 
This  relationship  is  true  only  if  negligible 
current  is  drawn  by  a load  on  the  output  ter- 
minals. 


HANDBOOK 


Resonant  Circuits  53 


J 

■ R 

r 

=:Xc  Ez 

_ L 

E.=  E,  . 

"Vrz+XC^ 


@ 


£i 


oJ 


_1^ 

T 


Xc 


-o 


-o 


o- 


Ei 


o 


E2=  Ei 


Xu 

Xl-Xc 


E2  = Ei 


XC 

'V'R2+  (Xl-Xc)2 


© 


E3  = Et  == 

KR2+(Xl-Xc)2 

E4  = Et 

KR^  + (Xi.-Xc)2 


Figure  17 

COMPLEX  A-C  VOLTAGE  DIVIDERS 


3-2  Resonant  Circuits 


A series  circuit  such  as  shown  in  figure  18 
is  said  to  be  in  resonance  when  the  applied 
frequency  is  such  that  the  capacitive  react- 
ance is  exactly  balanced  by  the  inductive  re- 
actance. At  this  frequency  the  two  reactances 
will  cancel  in  their  effects,  and  the  impedance 
of  the  circuit  will  be  at  a minimum  so  that 
maximum  current  will  flow.  In  fact,  as  shown 
in  figure  19  the  net  impedance  of  a series 
circuit  at  resonance  is  equal  to  the  resistance 
which  remains  in  the  circuit  after  the  react- 
ances have  been  cancelled. 

Resonant  Frequency  Some  resistance  is  always 
present  in  a circuit  be- 
cause it  is  possessed  in  some  degree  by  both 
the  inductor  and  the  capacitor.  If  the  fre- 
quency of  the  alternator  E is  varied  from 
nearly  zero  to  some  high  frequency,  there  will 
be  one  particular  frequency  at  which  the  in- 
ductive reactance  and  capacitive  reactance 
will  be  equal.  This  is  known  as  the  resonant 
frequency,  and  in  a series  circuit  it  is  the 
frequency  at  which  the  circuit  current  will  be 
a maximum.  Such  series  resonant  circuits  are 
chiefly  used  when  it  is  desirable  to  allow  a 
certain  frequency  to  pass  through  the  circuit 
(low  impedance  to  this  frequency),  while  at 
the  same  time  the  circuit  is  made  to  offer 
considerable  opposition  to  currents  of  other 
frequencies. 


Figure  18 

SERIES  RESONANT  CIRCUIT 


If  rhe  values  of  inductance  and  capacitance 
both  are  fixed,  there  will  be  only  one  resonant 
frequency. 

if  both  the  inductance  and  capacitance  are 
made  variable,  the  circuit  may  then  be  changed 
or  tuned,  so  that  a number  of  combinations 
of  inductance  and  capacitance  can  resonate  at 
the  same  frequency.  This  can  be  more  easily 
understood  when  one  considers  that  inductive 
reactance  and  capacitive  reactance  travel  in 
opposite  directions  as  the  frequency  is  changed. 
For  example,  if  the  frequency  were  to  remain 
constant  and  the  values  of  inductance  and 
capacitance  were  then  changed,  the  following 
combinations  would  have  equal  reactance: 

Frequency  is  constant  at  60  cycles. 


L is  expressed  in  henrys. 


C is  expressed  in 
farad.) 

L Xl 

.265  100 

2.65  1,000 

26.5  10,000 

265.00  100,000 

2,650.00  1,000,000 


microfarads 

(.000001 

C 

26.5 

100 

2.65 

1,000 

.265 

10,000 

.0265 

100,000 

.00265 

1,000,000 

Frequency  From  the  formula  for  reson- 

of  Resonance  ance,  2!rfL  = l/2!rfC,  the  res- 

onant frequency  is  determined: 


27rVLC 

where  f = frequency  in  cycles, 

L = inductance  in  henrys, 

C = capacitance  in  farads. 

It  is  more  convenient  to  express  L and  C 
in  smaller  units,  especially  in  making  radio- 
frequency calculations;  f can  also  be  ex- 
pressed in  megacycles  or  kilocycles.  A very 
useful  group  of  such  formulas  is; 

, 25,330  25,330  25,330 

f’  or  L = or  C = 

LC  f^C  f^L 

where  f = frequency  in  megacycles, 

L = inductance  in  microhenrys, 

C = capacitance  in  micromictofarads. 


54  Alternating  Current  Circuits 


THE  RADIO 


Figure  19 

IMPEDANCE  OF  A 
SERIES-RESONANT  CIRCUIT 

Sfcow/ng  the  vartatton  In  rmactancm  of  sopo- 
ror«  o/omonrs  and  In  tha  nat  Impadanea  of  a 
sortos  roMonant  circuit  (such  as  (Iguro  18)  with 
changing  froquoncy.  Tho  vortical  lino  Is  drawn 
at  tho  point  of  rosonanco  = 0)  In  tho 

sorlos  circuit. 


Impedance  of  Series  The  impedance  across 
Resonant  Circuits  the  terminals  of  a series 
resonant  circuit  (figure 

18)  is: 

z = 

where  Z = impedance  in  ohms, 
r = resistance  in  ohms, 

Xc  = capacitive  reactance  in  ohms, 

Xl  = inductive  reactance  in  ohms. 

From  this  equation,  it  can  be  seen  that  the 
impedance  is  equal  to  the  vector  sum  of  the 
circuit  resistance  and  the  difference  between 
the  two  reactances.  Since  at  the  resonant  fre- 
quency Xl  equals  Xc,  the  difference  between 
them  (figure  19)  is  zero,  so  that  at  resonance 
the  impedance  is  simply  equal  to  the  resist- 
ance of  the  circuit;  therefore,  because  the 
resistance  of  most  normal  radio-frequency 
circuits  is  of  a very  low  order,  the  impedance 
is  also  low. 

At  frequencies  higher  and  lower  than  the 
resonant  frequency,  the  difference  between 
the  reactances  will  be  a definite  quantity  and 
will  add  with  the  resistance  to  make  the  im- 
pedance higher  and  higher  as  the  circuit  is 
tuned  off  the  resonant  frequency. 

If  Xc  should  be  greater  than  Xl,  then  the 
term  (Xl  - Xc)  will  give  a negative  number. 
However,  when  the  difference  is  squared  the 
product  is  always  positive.  This  means  that 
the  smaller  reactance  is  subtracted  from  the 
larger,  regardless  of  whether  it  be  capacitive 
or  inductive,  and  the  difference  squared. 


Figura  20 

RESONANCE  CURVE 

Showing  tho  Ineroaso  In  Impodanco  at  roson- 
once  for  o poro//ef>resononf  circuit,  and  simi- 
larly, tho  Ineroaso  In  curront  at  rosonanco  for 
a sorlos-rosonant  circuit.  Tho  sharpnoss  of 
rosonanco  is  dotormlnod  hy  tho  Q of  tho  circuit, 
os  lllustratod  by  a comparison  botwoon  A, 
B,  and  C. 


Current  and  Voltoge 
in  Series  Resonant 
Circuits 

Ohm’s  law. 


Formulas  for  calculating 
currents  and  voltages  in 
a series  resonant  circuit 
are  similar  to  those  of 


E 

1.=—  E = IZ 
Z 


The  complete  equations: 

E 

I = 

Vr^  +(Xl  - Xc)* 


E = I Vr’+(XL  -Xc)^ 


Inspection  of  the  above  formulas  will  show 
the  following  to  apply  to  series  resonant  cir- 
cuits: When  the  impedance  is  low,  the  current 
will  be  high;  conversely,  when  the  impedance 
is  high,  the  current  will  be  low. 

Since  it  is  known  that  the  impedance  will 
be  very  low  at  the  resonant  frequency,  it  fol- 
lows that  the  current  will  be  a maximum  at 
this  point.  If  a graph  is  plotted  of  the  current 
against  the  frequency  either  side  of  resonance, 
the  resultant  curve  becomes  what  is  known  as 
a resonance  curve.  Such  a curve  is  shown  in 
figure  20,  the  frequency  being  plotted  against 
current  in  the  series  resonant  circuit. 

Several  factors  will  have  an  effect  on  the 
shape  of  this  resonance  curve,  of  which  re- 


HANDBOOK 


Circuit  Q 55 


sistance  and  L-to-C  ratio  are  the  important 
considerations.  The  cutves  B and  C in  figure 
20  show  the  effect  of  adding  increasing  values 
of  resistance  to  the  circuit.  It  will  be  seen 
that  the  peaks  become  less  and  less  prominent 
as  the  tesistance  is  increased;  thus,  it  can  be 
said  that  the  selectivity  of  the  circuit  is 
thereby  decreased.  Selectivity  in  this  case 
can  be  defined  as  the  ability  of  a circuit  to 
discriminate  against  frequencies  adjacent  to 
the  resonant  frequency. 

Voltoge  Across  Coil  Because  the  a.c.  or  r-f 
ond  Copocitor  in  voltage  across  a coil  and 

Series  Circuit  capacitor  is  proportional 

to  the  reactance  (for  a 
given  current),  the  actual  voltages  across  the 
coil  and  across  the  capacitor  may  be  many 
times  greater  than  the  terminal  voltage  of  the 
circuit.  At  resonance,  the  voltage  across  the 
coil  (or  the  capacitor)  is  Q times  the  applied 
voltage.  Since  the  Q (or  merit  factor)  of  a 
series  circuit  can  be  in  the  neighborhood  of 
100  or  more,  the  voltage  across  the  capacitor, 
for  example,  may  be  high  enough  to  cause 
flashover,  even  though  the  applied  voltage  is 
of  a value  considerably  below  that  at  which 
the  capacitor  is  rated. 

Circuit  Q — Sharp-  An  extremely  important 
n«ss  of  Resonance  ptoperty  of  a capacitor  or 
an  inductor  is  its  factor- 
of-merit,  more  generally  called  its  Q.  It  is  this 
factor,  Q,  which  primarily  determines  the 
sharpness  of  resonance  of  a tuned  circuit. 
This  factot  can  be  exptessed  as  the  ratio  of 
the  reactance  to  the  resistance,  as  follows: 


2nfL 


where  R = total  resistance. 

Skin  Effect  The  actual  resistance  in  a wire 
or  an  inductor  can  be  far  greater 
than  the  d-c  value  when  the  coil  is  used  in  a 
radio-frequency  circuit;  this  is  because  the 
current  does  not  travel  through  the  entire 
cross-section  of  the  conductor,  but  has  a tend- 
ency to  travel  closer  and  closer  to  the  surface 
of  the  wire  as  the  frequency  is  increased.  This 
is  known  as  the  skin  effect. 

The  actual  current-carrying  portion  of  the 
wire  is  decreased,  as  a result  of  the  skin 
effect,  so  that  the  ratio  of  a-c  to  d-c  tesist- 
ance of  the  wire,  called  the  resistance  ratio, 
is  increased.  The  resistance  tatio  of  wires  to 
be  used  at  frequencies  below  about  500  kc. 
may  be  materially  reduced  through  the  use  of 
litz  wite.Litz  wire,  of  the  type  commonly  used 
to  wind  the  coils  of  455-kc.  i-f  transformers, 
may  consist  of  3 to  10  strands  of  insulated 
wire,  about  No.  40  in  size,  with  the  individual 


strands  connected  together  only  at  the  ends  of 
the  coils. 

Variation  of  Q Examination  of  the  equation 
with  Frequency  for  determining  Q might  give 
rise  to  the  thought  that  even 
though  the  resistance  of  an  inductot  increases 
with  frequency,  the  inductive  reactance  does 
likewise,  so  that  the  Q might  be  a constant. 
Actually,  however,  it  works  out  in  practice 
that  the  Q of  an  inductor  will  reach  a relative- 
ly broad  maximum  at  some  particular  frequency. 
Hence,  coils  normally  are  designed  in  such  a 
manner  that  the  peak  in  their  curve  of  Q with 
frequency  will  occur  at  the  normal  operating 
frequency  of  the  coil  in  the  circuit  for  which 
it  is  designed. 

The  Q of  a capacitor  ordinarily  is  much 
higher  than  that  of  the  best  coil.  Therefore, 
it  usually  is  the  merit  of  the  coil  that  limits 
the  overall  Q of  the  circuit. 

At  audio  frequencies  the  core  losses  in  an 
iron-core  inductor  greatly  reduce  the  Q from 
the  value  that  would  be  obtained  simply  by 
dividing  the  reactance  by  the  tesistance.  Ob- 
viously the  core  losses  also  represent  circuit 
resistance,  just  as  though  the  loss  occurred 
in  the  wire  itself. 

Parollel  In  radio  circuits,  parallel  reson- 
Resonance  ance  (more  correctly  termed  anti- 
resonance) is  more  frequently 
encountered  than  series  resonance;  in  fact,  it 
is  the  basic  foundation  of  receiver  and  trans- 
mitter circuit  operation.  A circuit  is  shown  in 
figure  21. 

The  "Tank”  In  this  circuit,  as  contrasted  with 
Circuit  a circuit  for  series  resonance,  L 

(inductance)  and  C (capacitance) 
are  connected  in  parallel,  yet  the  combination 
can  be  considered  to  be  in  series  with  the 
remainder  of  the  circuit.  This  combination 
of  L and  C,  in  conjunction  with  R,  the  re- 
sistance which  is  principally  included  in  L,  is 
sometimes  called  a tank  circuit  because  it  ef- 
fectively functions  as  a storage  tank  when  in- 
corporated in  vacuum  tube  circuits. 

Contrasted  with  series  resonance,  there  are 
two  kinds  of  current  which  must  be  considered 
in  a parallel  resonant  circuit:  (1)  the  line  cur- 
rent, as  read  on  the  indicating  meter  M,,  (2) 
the  circulating  current  which  flows  within  the 
parallel  L-C-R  portion  of  the  circuit.  See 
figure  21. 

At  the  resonant  frequency,  the  line  current 
(as  read  on  the  meter  M,)  will  drop  to  a very 
low  value  although  the  circulating  current  in 
the  L-C  circuit  may  be  quite  large.  It  is  inter- 
esting to  note  that  the  parallel  resonant  cir- 
cuit acts  in  a distinctly  opposite  manner  to 
that  of  a series  resonant  circuit,  in  which  the 


56  Alternating  Current  Circuits 


THE  RADIO 


Figure  21 

PARALLEL-RESONANT  CIRCUIT 

The  inductance  L and  capacitance  C comprise 
the  reactive  elements  of  the  parallel-resonant 
(anti-resofiant)  tank  circuit,  and  the  resistance 
R indicates  the  sum  of  the  r-f  resistance  of  the 
coil  and  capacitor,  plus  the  resisfonce  coupled 
into  the  circuit  from  the  external  load.  In  most 
cases  the  tuning  capacitor  has  much  lower  r-f 
resistance  than  the  coll  and  can  therefore  be 
ignored  in  comparison  with  the  coil  resistance 
and  the  coupled-ln  resistance.  The  instrument 
Ml  indicotes  the  **//ne  current**  which  keeps 
the  circuit  in  a state  of  oscillation  — this  cur- 
rent is  the  seme  as  the  fundamental  component 
of  the  plate  current  of  a Class  C amplifier  which 
might  be  feeding  the  tank  circuit.  The  in- 
strument indicates  the  **tank  current**  which 
is  equal  to  the  line  current  multiplied  by  the 
operating  Q of  the  tank  circuit. 


current  is  at  a maximum  and  the  impedance  is 
minimum  at  resonance.  It  is  for  this  reason 
that  in  a parallel  resonant  circuit  the  principal 
consideration  is  one  of  impedance  rather  than 
current.  It  is  also  significant  that  the  imped- 
ance  curve  for  parallel  circuits  is  very  nearly 
identical  to  that  of  the  current  curve  for  series 
resonance.  The  impedance  at  resonance  is 
expressed  as: 

(277fL)' 

Z = , 

R 

where  Z = impedance  in  ohms, 

L = inductance  in  henrys, 
f = frequency  in  cycles, 

R = resistance  in  ohms. 

Or,  impedance  can  be  expressed  as  a func- 
tion of  Q as: 

Z = 27TfLQ, 

showing  th^t  the  impedance  of  a circuit  is 
directly  proportional  to  its  effective  Q at 
resonance. 

The  curves  illustrated  in  figure  20  can  be 
applied  to  parallel  resonance.  Reference  to  the 
curve  will  show  that  the  effect  of  adding  re- 
sistance to  the  circuit  will  result  in  both  a 
broadening  out  and  lowering  of  the  peak  of  the 
curve.  Since  the  voltage  of  the  circuit  is 
directly  proportional  to  the  impedance,  and 
since  it  is  this  voltage  that  is  applied  to  the 
grid  of  the  vacuum  tube  in  a detector  or  am- 


plifier circuit,  the  impedance  curve  must  have 

a sharp  peak  in  order  for  the  circuit  to  be 
selective.  If  the  curve  is  broad-topped  in 
shape,  both  the  desired  signal  and  the  inter- 
fering signals  at  close  proximity  to  resonance 
will  give  nearly  equal  voltages  on  the  grid  of 
the  tube,  and  the  circuit  will  then  be  non- 
selective;  i.e.,  it  will  tune  broadly. 

Effect  of  L/C  Ratio  In  order  that  the  highest 
in  Parallel  Circuits  possible  voltage  can  be 
developed  across  a paral- 
lel resonant  circuit,  the  impedance  of  this 
circuit  must  be  very  high.  The  impedance  will 
be  greater  with  conventional  coils  of  limited 
Q when  the  ratio  of  inductance-to- capacitance 
is  great,  that  is,  when  L is  large  as  compared 
with  C-  When  the  resistance  of  the  circuit  is 
very  low.  Xl  will  equal  Xq  at  maximum  im- 
pedance. There  are  innumerable  ratios  of  L 
and  C that  will  have  equal  reactance,  at  a 
given  resonant  frequency,  exactly  as  in  the 
case  in  a series  resonant  circuit. 

In  practice,  where  a certain  value  of  in- 
ductance is  tuned  by  a variable  capacitance 
over  a fairly  wide  range  in  frequency,  the 
L/C  ratio  will  be  small  at  the  lowest  fre- 
quency endandlargeatthe  high-frequency  end. 
The  circuit,  therefore,  will  have  unequi  gain 
and  selectivity  at  the  two  ends  of  the  band  of 
frequencies  which  is  being  tuned.  Increasing 
the  Q of  the  circuit  (lowering  the  resistance) 
will  obviously  increase  both  the  selectivity 
and  gain. 

Circulating  Tank  The  Q of  a circuit  has 

Current  at  Resonance  a definite  bearing  on 
the  circulating  tank 
current  at  resonance.  This  tank  current  is 
very  nearly  the  value  of  the  line  current  multi- 
plied by  the  effective  circuit  Q.  For  example; 
an  r-f  line  current  of  0.050  amperes,  with  a 
circuit  Q of  100,  will  give  a circulating  tank 
current  of  approximately  5 amperes.  From  this 
it  can  be  seen  that  both  the  inductor  and  the 
connecting  wires  in  a circuit  with  a high  Q 
must  be  of  very  low  resistance,  particularly  in 
the  case  of  high  power  transmitters,  if  heat 
losses  are  to  be  held  to  a minimum. 

Because  the  voltage  across  the  tank  at 
resonance  is  determined  by  the  Q,  it  is  pos- 
sible to  develop  very  high  peak  voltages 
across  a high  Q tank  with  but  little  line  cur- 
rent. 

Effect  of  Coupling  If  a parallel  resonant  cir- 
on  Impedance  cuit  is  coupled  to  another 

circuit,  such  as  an  antenna 
output  circuit,  the  impedance  and  the  effective 
Q of  the  parallel  circuit  is  decreased  as  the 
coupling  becomes  closer.  The  effect  of  closer 
(tighter)  coupling  is  the  same  as  though  an 


HANDBOOK 


Circuit  Impedance  57 


Figure  22 

EFFECT  OF  COUPLING  ON  CIRCUIT  IMPEDANCE  AND  Q 


actual  resistance  were  added  in  series  with 
the  parallel  tank  circuit.  The  resistance  thus 
coupled  into  the  tank  circuit  can  be  con- 
sidered as  being  reflected  from  the  output  or 
load  circuit  to  the  driver  circuit. 

The  behavior  of  coupled  circuits  depends 
largely  upon  the  amount  of  coupling,  as  shown 
in  figure  22.  The  coupled  current  in  the  sec- 
ondary circuit  is  small,  varying  with  frequency, 
being  maximum  at  the  resonant  frequency  of 
the  circuit.  As  the  coupling  is  increased 
between  the  two  circuits,  the  secondary  res- 
onance curve  becomes  broader  and  the  reso- 
nant amplitude  increases,  until  the  reflected 
resistance  is  equal  to  the  primary  resistance. 
This  point  is  called  the  critical  coupling 
point.  With  greater  coupling,  the  secondary 
resonance  curve  becomes  broader  and  develops 
double  resonance  humps,  which  become  more 
pronounced  and  farther  apart  in  frequency  as 
the  coupling  between  the  two  circuits  is 
increased. 

Tonk  Circuit  When  the  plate  circuit  of  a 
Flywheel  Effect  Class  B or  Class  C operated 
tube  is  connected  to  a par- 
allel resonant  circuit  tuned  to  the  same  fre- 
quency as  the  exciting  voltage  for  the  ampli- 
fier, the  plate  current  serves  to  maintain  this 
L/C  circuit  in  a state  of  oscillation. 

The  plate  currentis  supplied  in  short  pulses 
which  do  not  begin  to  resemble  a sine  wave, 
even  though  the  grid  may  be  excited  by  a sine- 
wave  voltage.  These  spurts  of  plate  current 
are  converted  into  a sine  wave  in  the  plate 
tank  circuit  by  virtue  of  the  "Q”  or  "flywheel 
effect”  of  the  tank. 

If  a tank  did  not  have  some  resistance 
losses,  it  would,  when  given  a "kick”  with  a 
single  pulse,  continue  to  oscillate  indefinitely. 
With  a moderate  amount  of  resistance  or  "fric- 
tion” in  the  circuit  the  tank  will  still  have 


inertia,  and  continue  to  oscillate  with  de- 
creasing amplitude  for  a time  after  being  given 
a "kick.”  With  such  a circuit,  almost  pure 
sine-wave  voltage  will  be  developed  across 
the  tank  circuit  even  though  power  is  supplied 
to  the  tank  in  short  pulses  or  spurts,  so  long 
as  the  spurts  ate  evenly  spaced  with  respect 
to  time  and  have  a frequency  that  is  the  same 
as  the  resonant  frequency  of  the  tank. 

Another  way  to  visualize  the  action  of  the 
tank  is  to  recall  that  a resonant  tank  with 
moderate  Q will  discriminate  strongly  against 
harmonics  of  the  resonant  frequency.  The  dis- 
torted plate  current  pulse  in  a Class  C ampli- 
fier contains  not  only  the  fundamental  fre- 
quency (that  of  the  grid  excitation  voltage) 
but  also  higher  harmonics.  As  the  tank  offers 
low  impedance  to  the  harmonics  and  high  im- 
pedance to  the  fundamental  (being  resonant  to 
the  latter),  only  the  fundamental  — a sine- 
wave  voltage  — appears  across  the  tank  circuit 
in  substantial  magnitude. 

Loaded  and  Confusion  sometimes  exists  as 
Unloaded  Q to  the  relationship  between  the 
unloaded  and  the  loaded  Q of  the 
tank  circuit  in  the  plate  of  an  r-f  power  ampli- 
fier. In  the  normal  case  the  loaded  Q of  the 
tank  circuit  is  determined  by  such  factors  as 
the  operating  conditions  of  the  amplifier,  band- 
width of  the  signal  to  be  emitted,  permissible 
level  of  harmonic  radiation,  and  such  factors. 
The  normal  value  of  loaded  Q for  an  r-f  ampli- 
fier used  for  communications  service  is  from 
perhaps  6 to  20.  The  unloaded  Q of  the  tank 
circuit  determines  the  efficiency  of  the  output 
circuit  and  is  determined  by  the  losses  in  the 
tank  coil,  its  leads  and  plugs  and  jacks  if  any, 
and  by  the  losses  in  the  tank  capacitor  which 
ordinarily  are  very  low.  The  unloaded  Q of  a 
good  quality  large  diameter  tank  coil  in  the 
high-frequency  range  may  be  as  high  as  500 


58  Alternating  Current  Circuits 


THE  RADIO 


to  800,  and  values  greater  than  300  are  quite 

common. 

Tank  Circuit  Since  the  unloaded  Q of  a tank 
Efficiency  circuit  is  determined  by  the 
minimum  losses  in  the  tank, 
while  the  loaded  Q is  determined  by  useful 
loading  of  the  tank  circuit  from  the  external 
load  in  addition  to  the  internal  losses  in  the 
tank  circuit,  the  relationship  between  the  two 
Q values  determines  the  operating  efficiency 
of  the  tank  circuit.  Expressed  in  the  form  of 
an  equation,  the  loaded  efficiency  of  a tank 
circuit  is: 

Tank  efficiency  = 1 x 100 

Qu 

where  Qu  = unloaded  Q of  the  tank  circuit 
Ql  = loaded  Q of  the  tank  circuit 
As  an  example,  if  the  unloaded  Q of  the 
tank  circuit  for  a class  C r-f  power  amplifier 
is  400,  and  the  external  load  is  coupled  to  the 
tank  circuit  by  an  amount  such  that  the  loaded 
Q is  20,  the  tank  circuit  efficiency  will  be: 
eff.  = (1  - 20/400)  X 100,  or  (1  - 0.05)  x 100, 
or  95  per  cent.  Hence  5 pet  cent  of  the  power 
output  of  the  Class  C amplifier  will  be  lost 
as  heat  in  the  tank  circuit  and  the  remaining 
95  per  cent  will  be  delivered  to  the  load. 

3-3  Nonsinusoidal  Waves 

and  Transients 

Pure  sine  waves,  discussed  previously,  are 
basic  wave  shapes.  Waves  of  many  different 
and  complex  shape  ate  used  in  electronics, 
particularly  square  waves,  saw-tooth  waves, 
and  peaked  waves. 

Wave  Composition  Any  periodic  wave  (one  that 
repeats  itself  in  definite 
time  intervals)  is  composed  of  sine  waves  of 
different  frequencies  and  amplitudes,  added 
together.  The  sine  wave  which  has  the  same 
frequency  as  the  complex,  periodic  wave  is 
called  the  fundamental.  The  frequencies  higher 
than  the  fundamental  are  called  harmonics, 
and  are  always  a whole  number  of  times  higher 
than  the  fundamental.  For  example,  the  fre- 
quency twice  as  high  as  the  fundamental  is 
called  the  second  harmonic. 

The  Square  Wave  Figure  23  compares  a square 
wave  with  a sine  wave  (A) 
of  the  same  frequency.  If  another  sine  wave 
(B)  of  smaller  amplitude,  but  three  times  the 
frequency  of  (A),  called  the  third  harmonic,  is 
added  to  (A),  the  resultant  wave  (C)  mote 
nearly  approaches  the  desired  square  wave. 


COMPOSITE  WAVE-FUNDAMENTAL 
PLUS  THIRD  HARMONIC 


Figure  24 

THIRD  HARMONIC  WAVE  PLUS 
FIFTH  HARMONIC 


Figure  25 

RESULTANT  WAVE,  COMPOSED  OF 
FUNDAMENTAL,  THIRD,  FIFTH, 
AND  SEVENTH  HARMONICS 


This  resultant  curve  (figure  24)  is  added  to 
a fifth  harmonic  curve  (D),  and  the  sides  of 
the  resulting  curve  (E)  ate  steeper  than  before. 
This  new  curve  is  shown  in  figure  25  after  a 
7th  harmonic  component  has  been  added  to  it, 
making  the  sides  of  the  composite  wave  even 
steeper.  Addition  of  mote  higher  odd  harmonics 
will  bring  the  resultant  wave  nearer  and  nearer 
to  the  desired  square  wave  shape.  The  square 
wave  will  be  achieved  if  an  infinite  number  of 
odd  harmonics  are  added  to  the  original  sine 
wave. 


HANDBOOK 


N 0 n s i n u so i d a I Waves  59 


Figure  26 

COMPOSITION  OF  A SAWTOOTH  WAVE 


The  Sawtooth  Wove  In  the  same  fashion,  a 
sawtooth  wave  is  made  up 
of  different  sine  waves  (figure  26).  The  ad- 
dition of  all  harmonics,  odd  and  even,  produces 
the  sawtooth  wave  form. 

The  Peaked  Wove  Figure  27  shows  the  com- 
position of  a peaked  wave. 
Note  how  the  addition  of  each  sucessive  har- 
monic makes  the  peak  of  the  resultant  higher 
and  the  sides  steeper. 

Other  Waveforms  The  three  preceeding  ex- 
amples show  how  a complex 
periodic  wave  is  composed  of  a fundamental 
wave  and  different  harmonics.  The  shape  of 
the  resultant  wave  depends  upon  the  harmonics 
that  are  added,  their  relative  amplitudes,  and 
relative  phase  relationships.  In  general,  the 
steeper  the  sides  of  the  waveform,  the  mote 
harmonics  it  contains. 

AC  Transient  Circuits  If  an  a-c  voltage  is  sub- 
stituted for  the  d-c  in- 
put voltage  in  the  RC  Transient  circuits  dis- 
cussed in  Chapter  2,  the  same  principles  may 
be  applied  in  the  analysis  of  the  transient 
behavior.  An  RC  coupling  circuit  is  designed 
to  have  a long  time  constant  with  respect  to 
the  lowest  frequency  it  must  pass.  Such  a 
circuit  is  shown  in  figure  28.  If  a nonsinus- 
oidal  voltage  is  to  be  passed  unchanged 
through  the  coupling  circuit,  the  time  constant 


© 


Figure  27 

COMPOSITION  OF  A PEAKED  WAVE 


must  be  long  with  respect  to  the  period  of  the 
lowest  frequency  contained  in  the  voltage 
wave. 

RC  Differentiator  An  RC  voltage  divider  that 
and  Integrator  is  designed  to  distort  the  in- 
put waveform  is  known  as  a 
differentiator  or  integrator,  depending  upon 
the  locations  of  the  output  taps.  The  output 
from  a differentiator  is  taken  across  the  re- 
sistance, while  the  output  from  an  integrator 
is  taken  across  the  capacitor.  Such  circuits 
will  change  the  shape  of  any  complex  a-c 
waveform  that  is  impressed  upon  them.  This 
distortion  is  a function  of  the  value  of  the 
time  constant  of  the  circuit  as  compared  to 
the  period  of  the  waveform.  Neither  a dif- 
ferentiator nor  an  integrator  can  change  the 


60  Alternating  Current  Circuits 


THE  RADIO 


PERIOD  OF  e=  lOOOUSECONOS 


©c*  INTEGRATOR  OUTPUT 

eft  = OIFFERENTIATOR  OUTPUT 


Figure  28 

R-C  COUPLING  CIRCUIT  WITH 
LONG  TIME  CONSTANT 


Figure  29 

R-C  DIFFERENTIATOR  AND 
INTEGRATOR  ACTION  ON 
A SINE  WAVE 

shape  of  a pure  sine  wave,  they  will  merely 
shift  the  phase  of  the  wave  (figure  29)-  The 
differentiator  output  is  a sine  wave  leading 
the  input  wave,  and  the  integrator  output  is  a 
sine  wave  which  lags  the  input  wave.  The  sum 
of  the  two  outputs  at  any  instant  equals  the 
instantaneous  input  voltage. 

Square  Wave  Input  If  a square  wave  voltage  is 
impressed  on  the  circuit  of 
figure  30,  a square  wave  voltage  output  may 
be  obtained  across  the  integrating  capacitor 
if  the  time  constant  of  the  circuit  allows  the 
capacitor  to  become  fully  charged.  In  this 
particular  case,  the  capacitor  never  fully 
charges,  and  as  a result  the  output  of  the 
integrator  has  a smaller  amplitude  than  the 
input.  The  differentiator  output  has  a maximum 
value  greater  than  the  input  amplitude,  since 
the  voltage  left  on  the  capacitor  from  the 
previous  half  wave  will  add  to  the  input  volt- 
age. Such  a circuit,  when  used  as  a differen- 
tiator, is  often  called  a peaker.  Peaks  of 
twice  the  input  amplitude  may  be  produced. 


Figure  30 

R-C  DIFFERENTIATOR  AND 
INTEGRATOR  ACTION  ON 
A SQUARE  WAVE 


Sawtooth  Wove  Input  If  a back-to-back  saw- 
tooth voltage  is  applied 
to  an  RC  circuit  having  a time  constant  one- 
sixth  the  period  of  the  input  voltage,  the  re- 
sult is  shown  in  figure  31-  The  capacitor 
voltage  will  closely  follow  the  input  voltage, 
if  the  time  constant  is  short,  and  the  integra- 
tor output  closely  resembles  the  input.  The 
an^litude  is  slightly  reduced  and  there  is  a 
slight  phase  lag.  Since  the  voltage  across  the 
capacitor  is  increasing  at  a constant  rate,  the 
charging  and  discharging  current  is  constant. 
The  output  voltage  of  the  differentiator,  there- 
fore, is  constant  during  each  half  of  the  saw- 
tooth input. 

Miscellaneous  Various  voltage  waveforms 
Inputs  other  than  those  represented 

here  may  be  applied  to  short 
RC  circuits  for  the  purpose  of  producing 
across  the  resistor  an  output  voltage  with  an 
amplitude  proportional  to  the  rate  of  change  of 
the  input  signal.  The  shorter  the  RC  time  con- 
stant is  made  with  respect  to  the  period  of  the 
input  wave,  the  mote  nearly  the  voltage  across 


HANDBOOK 


T ransformers  61 


e = 100  V, 
(peak) 
1000  C.p.s. 


INTEGRATOR 
OUTPUT  (^c ) 

DIFFERENTIATOR 
OUTPUT  (Cr) 


+10 

eo  — Z'- V / OUTPUT  OF  OIFF- 

° ^ / EREgTIATOR(eR) 


Figure  31 

R-C  DIFFERENTIATOR  AND 
INTEGRATOR  ACTION  ON 
A SAWTOOTH  WAVE 


the  capacitor  conforms  to  the  input  voltage. 
Thus,  the  differentiator  output  becomes  of 
particular  importance  in  very  short  RC  cir- 
cuits. Differentiator  outputs  for  various  types 
of  input  waves  are  shown  in  figure  32. 

Square  Wave  Test  The  application  of  a square 
for  Audio  Equipment  wave  input  signal  to  audio 
equipment,  and  the  ob- 
servation of  the  reproduced  output  signal  on 
an  oscilloscope  will  provide  a quick  and  ac- 
curate check  of  the  overall  operation  of  audio 
equipment.  Low-frequency  and  high-frequency 
response,  as  well  as  transient  response  can  be 
examined  easily.  If  the  amplifier  is  deficient 
in  low-frequency  response,  the  flat  top  of  the 
square  wave  will  be  canted,  as  in  figure  33- 
If  the  high-frequency  response  is  inferior,  the 
tise  time  of  the  output  wave  will  be  retarded 
(figure  34).  An  amplifier  with  a limited  high- 
and  low-  frequency  response  will  turn  the 
square  wave  into  the  approximation  of  a saw- 
tooth wave  (figure  (35). 

3-4  Transformers 

When  two  coils  are  placed  in  such  inductive 
relation  to  each  other  that  the  lines  of  force 
from  one  cut  across  the  turns  of  the  other 
inducing  a current,  the  combination  ctui  be 
called  a transformer.  The  name  is  derived  from 
the  fact  that  energy  is  transformed  from  one 
winding  to  another.  The  inductance  in  which 


Dlff^rmntlator  outputs  of  short  r-e  c/rcutts  for 
various  Input  voltage  waveshapes.  The  output 
voltage  Is  proportional  to  the  rate  of  change 
of  the  Input  voltage. 


the  original  flux  is  produced  is  called  the 
primary;  the  inductance  which  receives  the 
induced  current  is  called  the  secondary.  In  a 
radio  receiver  power  transformer,  for  example, 
the  coil  through  which  the  110-volt  a.c.  passes 
is  the  primary,  and  the  coil  from  which  a higher 
or  lower  voltage  than  the  a-c  line  potential  is 
obtained  is  the  secondary. 

Transformers  can  have  either  air  or  mag- 
netic cores,  depending  upon  the  frequencies  at 
which  they  are  to  be  operated.  The  reader 
should  thoroughly  impress  upon  his  mind  the 
fact  that  current  can  be  transferred  from  one 
circuit  to  another  only  if  the  primary  current 
is  changing  or  alternating.  From  this  it  can  be 
seen  that  a power  transformer  cannot  possibly 
function  as  such  when  the  primary  is  supplied 
with  non-pulsating  d.c. 

A power  transformer  usually  has  a magnetic 
core  which  consists  of  laminations  of  iron, 
built  up  into  a square  or  rectangular  form, 
with  a center  opening  or  window.  The  second- 
ary windings  may  be  several  in  number,  each 
perhaps  delivering  a diffetent  voltage.  The 
secondary  voltages  will  be  proportional  to  the 
turns  ratio  and  the  primary  voltage. 


62  Alternating  Current  Circuits 


THE  RADIO 


Figure  33 

Amplifi»r  dmftcl^nt  In  low  froquoncy  romponso  will  tilstort  aquaro  wove  appllod  to  tho  Input  circuit,  aa 
ahown.  A M^cyclo  aquato  wav  may  ho  uaod. 

A:  Drop  In  gain  of  /ow  froqifeneios 

B?  l.*o<//ng  pitoso  ahllt  at  low  froquoncloa 
C:  Lagging  pfioso  sfi//f  of  low  froquoncloa 

D:  Acconfuatod  low  froquoncy  gain 


Types  of  Transformers  are  used  in  alter- 
Transformers  nating-current  circuits  to  trans- 
fer power  at  one  voltage  and  im- 
pedance to  another  circuit  at  another  voltage 
and  impedance.  There  are  three  main  classifi- 
cations of  transformers:  those  made  for  use 
in  power-frequency  circuits,  those  made  for 
audio-frequency  applications,  and  those  made 
for  radio  frequencies. 

Tho  Tronsformotlon  In  a perfect  transformer  all 
Ratio  the  magnetic  flux  lines 

produced  by  rhe  primary 
winding  link  every  turn  of  the  secondary  wind- 
ing. For  such  a transformer,  the  ratio  of  the 
primary  and  secondary  voltages  is  exactly 
the  same  as  the  ratio  of  the  number  of  turns 
in  the  two  windings: 

Np  Ep 

where  Np  = number  of  turns  in  the  primary 
winding 

Ns  = number  of  turns  in  the  secondary 
winding 

Ep  = voltage  across  the  primary  winding 


Es  = voltage  across  the  secondary 
winding 

In  pracrice,  rhe  transformation  ratio  of  a 
transformer  is  somewhat  less  than  the  turns 
ratio,  since  unity  coupling  does  not  exist 
between  the  primary  and  secondary  windings. 

Ampere  Turn*  (Nl)  The  current  that  flows  in 
the  secondary  winding  as  a 
result  of  the  induced  voltage  must  produce  a 
flux  which  exactly  equals  the  primary  flux. 
The  magnetizing  force  of  a coil  is  expressed 
as  the  product  of  the  number  of  turns  in  the 
coil  times  the  current  flowing  in  it: 

Np  Is 

NpXlp=NsxIs)  = — 

Ns  Ip 

where  Ip  = primary  current 

Is  = secondary  current 

It  can  be  seen  from  this  expression  that 
when  the  voltage  is  stepped  up,  the  current 
is  stepped  down,  and  vice-versa. 

Leakoge  Reactance  Since  unity  coupling  does 
not  exist  in  a practical 


Figure  34 

Output  wavmuhapu  of  ampllfit  having  daftelancy 
In  hlgh’frmguaney  response.  Tamtad  with  10-ke. 
squors  wavs. 


Figure  35 


Output  wovaahapa  of  ampllflar  having  llmltad 
low-fraquancy  and  high-fraguancy  response. 
Tautad  with  I-kc.  squore  wove. 


HANDBOOK 


Electric  Filters  63 


+ 8 


Figure  36 

IMPEDANCE-MATCHING  TRANSFORMER 

T/ie  ref/ecre(/  /mpec/ance  vorfes  dir*city  In 
proportion  to  tho  socondary  lood  and 

diroctly  In  proportion  to  tho  aquaro  of  tho 
primary-to-aoeondary  turna  ratio. 


transformer,  part  of  the  flux  passing  from  the 
primary  circuit  to  the  secondary  circuir  fol- 
lows a magnetic  circuit  acted  upon  by  the 
primary  only.  The  same  is  true  of  the  second- 
ary flux.  These  leakage  fluxes  cause  leakage 
reactance  in  the  transformer,  and  tend  to 
cause  the  traitsformer  to  have  poor  voltage 
regulation.  To  reduce  such  leakage  reactance, 
the  primary  and  secondary  windings  should 
be  in  close  proximity  to  each  other.  The  more 
expensive  transformers  have  interleaved  wind- 
ings to  reduce  inherent  leakage  reactance. 

Impedance  In  the  ideal  transformer,  the 

Transformation  impedance  of  the  secondary 
load  is  reflected  back  into  the 
primary  winding  in  the  following  relationship: 

Zp  = N'Zs  , or  N = VZp/Zs 

where  Zp  = reflected  primary  impedance 
N = turns  ratio  of  transformer 
Zg  = impedance  of  secondary  load 

Thus  any  specific  load  connected  to  the 
secondary  terminals  of  the  transformer  will 
be  transformed  to  a different  specific  value 
appearing  across  the  primary  terminals  of  the 
transformer.  By  the  proper  choice  of  turns 
ratio,  any  reasonable  value  of  secondary  load 
impedance  may  be  "reflected”  into  the  pri- 
mary winding  of  the  transformer  to  produce  the 
desired  transformer  primary  impedance.  The 
phase  angle  of  the  primary  "reflected”  im- 
pedance will  be  the  same  as  the  phase  angle 
of  the  load  impedance.  A capacitive  second- 
ary load  will  be  presented  to  the  transformer 
source  as  a capacity,  a resistive  load  will 
present  a resistive  "reflection”  to  the  primary 
source.  Thus  the  primary  source  "sees”  a 
transformer  load  entirely  dependent  upon  the 
secondary  load  impedance  and  the  turns  ratio 
of  the  transformer  (figure  36). 

The  Auto  The  type  of  transformer  in  figure 
Transformer  37,  when  wound  with  heavy  wire 
over  an  iron  core,  is  a common 
device  in  primary  power  circuits  for  the  pur- 
pose of  increasing  or  decreasing  the  line  volt- 


Schomatle  diagram  of  an  auto~tranaformor 
ahowing  tho  mothod  of  connocting  It  to  tho  lino 
and  to  tho  load,  Whon  only  a amall  amount  of 
atop  up  or  atop  down  la  roquirod,  tho  auto- 
tranaformor  may  bo  much  amollor  phyalcally 
than  would  bo  a tranaformor  with  a aeparato 
aocondary  winding.  Continuoualy  varlablo 
outO“tranaformora  (Vartac  and  Poworatat)  ore 
widoly  uaod  cammorclally. 


age.  In  effect,  it  is  merely  a continuous  wind- 
ing with  taps  taken  at  various  points  along 
the  winding,  the  input  voltage  being  applied 
to  the  bottom  and  also  to  one  tap  on  the  wind- 
ing. If  the  output  is  taken  from  this  same 
tap,  the  voltage  ratio  will  be  1-to-l;  i.e.,  the 
input  voltage  will  be  the  same  as  the  output 
voltage.  On  the  other  hand,  if  the  output  tap 
is  moved  down  toward  the  common  terminal, 
there  will  be  a step-down  in  tbe  turns  ratio 
with  a consequent  step-down  in  voltage.  The 
initial  setting  of  the  middle  input  tap  is  chosen 
so  that  the  number  of  turns  will  have  suffi- 
cient reactance  to  keep  the  no-load  primary 
current  at  a reasonably  low  value. 

3-5  Electric  Filters 

There  are  many  applications  where  it  is 
desirable  to  pass  a d-c  component  without 
passing  a superimposed  a-c  component,  or  to 


ELEMENTARY  FILTER  SECTIONS 


L-5ECTION5  T-NETWORK 


Figure  38 

Comp/ex  fllfrs  moy  mo(/«  up  from  rfioso  baste 
f/ifr  saetlens. 


64  Alternating  Current  Circuits 


THE  RADIO 


LOW-PASS  SHUNT-DERIVED  FILTER 

(^SERIES-ARM  resonated) 


FREQUENCY 


HIGH-PASS  SERIES-DERIVED  FILTER 

(S><UNT-ARM  resonated) 


FREQUENCY 


R=  LOAD  RESISTANCE 


R = LOAD  RESISTANCE 
-Ck 
M ... 


L2>  L)^ 

M 


‘-''“Tfe 


fas  CUT-OFF  FREQUENCY,  /<  = FREQUENCY  OF 

HIGH  ATTENUATION 


/“l»CUT-OFF  FREQUENCY.  = FREQUENCY  OF 

HIGH  ATTENUATION 


Figure  39 

TYPICAL  LOW-PASS  AND  HIGH-PASS  FILTERS,  ILLUSTRATING  SHUNT  AND  SERIES 

DERIVATIONS 


pass  all  ftequeacies  above  or  below  a certaio 
frequency  while  rejecting  or  attenuating  all 
others,  or  to  pass  only  a certain  band  or  bands 
of  frequencies  while  attenuating  all  others. 

All  of  these  things  can  be  done  by  suitable 
combinations  of  inductance,  capacitance  and 
resistance.  However,  as  whole  books  have 
been  devoted  to  nothing  but  electric  filters,  it 
can  be  appreciated  that  it  is  possible  only  to 
touch  upon  them  superficially  in  a general 
coverage  book. 

Filter  Operation  A filter  acts  by  virtue  of  its 
property  of  offering  very  high 
impedance  to  the  undesired  frequencies,  while 
offering  but  little  impedance  to  the  desired 
frequencies.  This  will  also  apply  to  d.c.  with 
a superimposed  a-c  component,  as  d.c.  can 
be  considered  as  an  alternating  current  of  zero 
frequency  so  far  as  filter  discussion  goes. 

Basic  Filters  Filters  are  divided  into  four 
classes,  descriptive  of  the  fre- 
quency bands  which  they  are  designed  to 
transmit:  high  pass,  low  pass,  band  pass  and 
band  elimination.  Each  of  these  classes  of 
filters  is  made  up  of  elementary  filter  sections 
called  L sections  which  consist  of  a series 
element  (Za)  and  a parallel  element  (Zb)  as 


illustrated  in  figure  38.  A finite  number  of  L 
sections  may  be  combined  into  basic  filter 
sections,  called  T networks  or  pi  networks, 
also  shown  in  figure  38.  Both  the  T and  pi 
networks  may  be  divided  in  two  to  form  half- 
sections. 


Filter  Sections  The  most  common  filter  sec- 
tion is  one  in  which  the  two 
impedances  Za  and  Zg  ate  so  related  that 
their  arithmetical  product  is  a constant:  Za  x 
Zb  = K’  at  all  frequencies.  This  type  of  filter 
section  is  called  a constant-K  section. 

A section  having  a sharper  cutoff  frequency 
than  a constant-K  section,  but  less  attenua- 
tion at  frequencies  far  removed  from  cutoff  is 
the  M-derived  section,  so  called  because  the 
shunt  or  series  elfement  is  resonated  with  a 
reactance  of  the  opposite  sign.  If  the  comple- 
mentary reactance  is  added  to  the  series  atm, 
the  section  is  said  to  be  shunt  derived;  if 
added  to  the  shunt  arm,  series  derived.  Each 
impedance  of  the  M-detived  section  is  related 
to  a corresponding  impedance  in  the  constant- 
K section  by  some  factor  which  is  a function 
of  the  constant  m.  M,  in  turn,  is  a function  of 
the  ratio  between  the  cutoff  frequency  and 
the  frequency  of  infinite  attenuation,  and  will 


HANDBOOK 


Filter  Design  65 


Through  the  use  of  the  curves  and  eguatlons  which  accompany  the  diagrams  in  the  illustration  above  It  is 
possible  to  determine  the  correct  values  of  Inductance  and  copoc/fonce  for  the  usual  types  of  pl-section 

filters. 


have  some  value  between  zero  and  one.  As  the 
value  of  m approaches  zero,  the  sharpness  of 
cutoff  increases,  but  the  less  will  be  the 
attenuation  at  several  times  cutoff  fre<|uency. 
A value  of  0.6  may  be  used  for  min  most  appli- 
cations. The  *'notch”  frequency  is  determined 
by  the  resonant  frequency  of  the  tuned  filter 
element.  The  amount  of  attenuation  obtained 
at  the  **notch*'  when  a derived  section  is  used 
is  determined  by  the  effective  Q of  the  reso- 
nant arm  (figure  39)- 

Filter  Assembly  Constant-K  sections  and  de- 
rived sections  may  be  cas- 
caded to  obtain  the  combined  characteristics 
of  sharp  cutoff  and  good  remote  frequency 
attenuation.  Such  a filter  is  known  as  a cowz- 
posite  filter.  The  amount  of  attenuation  will 
depend  upon  the  number  of  filter  sections 
used,  and  the  shape  of  the  transmission  curve 
depends  upon  the  type  of  filter  sections  used. 
All  filters  have  some  insertion  loss.  This 
attenuation  is  usually  uniform  to  all  frequen- 


cies within  the  pass  band.  The  insertion  loss 
varies  with  the  type  of  filter,  the  Q of  the 
components  and  the  type  of  termination  em- 
ployed. 

Electric  Filter  Electric  wave  filters  have  long 
Design  been  used  in  some  amateur  sta- 

tions in  the  audio  channel  to 
reduce  the  transmission  of  unwanted  high  fre- 
quencies and  hence  to  reduce  the  bandwidth 
occupied  by  a radiophone  signal.  The  effec- 
tiveness of  a properly  designed  and  properly 
used  filter  circuit  in  reducing  QRM  and  side- 
band splatter  should  not  be  underestimated. 

In  recent  years,  high  frequency  filters  have 
become  commonplace  in  TVI  reduction.  High- 
pass  type  filters  are  placed  before  the  input 
stage  of  television  receivers  to  reject  the 
fundamental  signal  of  low  frequency  trans- 
mitters. Low-pass  filters  are  used  in  the  out- 
put circuits  of  low  frequency  transmitters  to 
prevent  harmonics  of  the  transmitter  from 
being  radiated  in  the  television  channels. 


66  Alternating  Current  Circuits 


The  chart  of  figure  40  gives  design  data 
and  procedure  on  the  pi-section  type  of  filter. 
M-derived  sections  with  an  M of  0.6  will  be 
found  to  be  most  satisfactory  as  the  input 
section  (or  half-section)  of  the  usual  filter 
since  the  input  impedance  of  such  a section 


is  most  constant  over  the  pass  band  of  the 

filter  section. 

Simple  filters  may  use  either  L,  T,  or  n sec- 
tions. Since  the  n section  is  the  more  com- 
monly used  type  figure  40  gives  design  data 
and  characteristics  for  this  type  of  filter. 


A PUSH-PULL  250-TH  AMPLIFIER  WITH  TVI  SHIELD  REMOVED 

Use  of  harmonic  filters  in  power  leads  and  antenna  circuit  reduces  radiation  of  TYf-producing  harmonics 
of  typical  push-pull  amplifier.  Shielded  enclosure  completes  harmonic  reduction  measures. 


CHAPTER  FOUR 


Vacuum  Tube  Principles 


In  the  previous  chapters  we  have  seen  the 
manner  in  which  an  electric  current  flows 
through  a metallic  conductor  as  a result  of  an 
electron  drift.  This  drift,  which  takes  place 
when  there  is  a difference  in  potential  between 
the  ends  of  the  metallic  conductor,  is  in  addi- 
tion to  the  normal  random  electron  motion 
between  the  molecules  of  the  conductor. 

The  electron  may  be  considered  as  a minute 
negatively  charged  particle,  having  a mass  of 
9 X 10~”  gram,  and  a charge  of  1.59  x 10“*’ 
coulomb.  Electrons  ace  always  identical, 
regardless  of  the  source  from  which  they  ate 
obtained. 

An  electric  current  can  be  caused  to  flow 
through  other  media  than  a metallic  conductor. 
One  such  medium  is  an  ionized  solution,  such 
as  the  sulfuric  acid  electrolyte  in  a storage 
battery.  This  type  of  current  flow  is  called 
electrolytic  conduction.  Further,  it  was  shown 
at  about  the  turn  of  the  century  that  an  elec- 
tric current  can  be  carried  by  a stream  of  free 
electrons  in  an  evacuated  chamber.  The  flow 
of  a current  in  such  a manner  is  said  to  take 
place  by  electronic  conduction.  The  study  of 
electron  tubes  (also  called  vacuum  tubes,  or 
valves)  is  actually  the  study  of  the  control  and 
use  of  electronic  currents  within  an  evacuated 
or  partially  evacuated  chamber. 

Since  the  current  flow  in  an  electron  tube 
takes  place  in  an  evacuated  chambet,  there 
must  be  located  within  the  enclosure  both  a 
source  of  electrons  and  a collector  for  the 
electrons  which  have  been  emitted.  The  elec- 
tron source  is  called  the  cathode,  and  the 
electron  collector  is  usually  called  the  anode. 
Some  external  source  of  energy  must  be  ap- 
plied to  the  cathode  in  order  to  impart  suffi- 
cient velocity  to  the  electrons  within  the 
cathode  material  to  enable  them  to  overcome 
the  surface  forces  and  thus  escape  into  the 
surrounding  medium.  In  the  usual  types  of 


electron  tubes  the  cathode  energy  is  applied 
in  the  form  of  heat;  electron  emission  from  a 
heated  cathode  is  called  thermionic  emission. 
In  another  common  type  of  electron  tube,  the 
photoelectric  cell,  energy  in  the  form  of  light 
is  applied  to  the  cathode  to  cause  photo- 
electric emission. 

4-'|  Thermionic  Emission 

Electron  Emission  of  electrons  from  the 
Emission  cathode  of  a thermionic  electron 
tube  takes  place  when  the  cathode 
of  the  tube  is  heated  to  a temperature  suffi- 
ciently high  that  the  free  electrons  in  the 
emitter  have  sufficient  velocity  to  overcome 
the  restraining  forces  at  the  surface  of  the 
material.  These  surface  forces  vary  greatly 
with  different  materials.  Hence  different  types 
of  cathodes  must  be  raised  to  different  temper- 
atures to  obtain  adequate  quantities  of  elec- 
tron emission.  The  several  types  of  emitters 
found  in  common  types  of  transmitting  and 
receiving  tubes  will  be  described  in  the  fol- 
lowing paragraphs. 

Cathode  Typos  The  emitters  or  cathodes  as 
used  in  present-day  thermi- 
onic electron  tubes  may  be  classified  into 
two  groups:  the  directly-heated  ot  filament 
iypeand  the  inditectly-heated  or  heater-cathode 
type.  Directly-heated  emitters  may  be  further 
subdivided  into  thtee  important  groups,  all 
of  which  are  commonly  used  in  modern  vacuum 
tubes.  These  classifications  are:  the  pure- 
tungsten  filament,  the  thoriated-tungsten 
filament,  and  the  oxide-coated  filament. 

The  Pure  Tung-  Pure  tungsten  wire  was  used 
sten  Filament  as  the  filament  in  nearly  all 
the  earlier  transmitting  and 


67 


68  Vacuum  Tube  Principles 


THE  RADIO 


Figure  1 

ELECTRON  TUBE  TYPES 


The  new  General  Electric  ceramic  triode  (6BY4)  is  shown  alongside  a con- 
ventional miniature  tube  (6265)  and  an  octal-based  receiving  tube  (25L6).  The  ceramic 
tube  is  designed  for  rugged  service  and  features  extremely  low  lead  inductance. 


receiving  tubes.  However,  the  thermionic  effi- 
ciency of  tungsten  wire  as  an  emitter  (the 
number  of  milliamperes  emission  per  watt  of 
filament  heating  power)  is  quite  low,  the  fila- 
ments become  fragile  after  use,  their  life  is 
rather  short,  and  they  are  susceptible  to  burn- 
out at  any  time.  Pure  tungsten  filaments  must 
be  tun  at  bright  white  heat  (about  2500°  Kel- 
vin). For  these  reasons,  tungsten  filaments 
have  been  replaced  in  all  applications  where 
another  type  of  filament  could  be  used.  They 
ate,  however,  still  universally  employed  in 
large  water-cooled  tubes  and  in  certain  large, 
high-power  air-cooled  triodes  where  another 
filament  type  would  be  unsuitable.  Tungsten 
filaments  are  the  most  satisfactory  for  high- 
power,  high-voltage  tubes  where  the  emitter 
is  subjected  to  positive  ion  bombardment 
caused  by  the  residual  gas  content  of  the 
tubes.  Tungsten  is  not  adversely  affected  by 
sucb  bombardment. 

The  Thoriated-  In  the  course  of  expeti- 
Tungsten  Filament  ments  made  upon  tungsten 
emitters,  it  was  found  that 
filaments  made  from  tungsten  having  a small 
amount  of  thoria  (thorium  oxide)  as  an  im- 
purity had  much  greater  emission  than  those 
made  from  the  pure  metal.  Subsequent  develop- 
ment has  resulted  in  the  highly  efficient  car- 
burized thoriated-tungsten  filament  as  used  in 
virtually  all  medium-power  transmitting  tubes 
today. 

Thoriated-tungsten  emitters  consist  of  a 
tungsten  wire  containing  from  1%  to  2%  thoria. 
The  activation  process  varies  between  dif- 
ferent manufacturers  of  vacuum  tubes,  but 
it  is  essentially  as  follows:  (1)  the  tube  is 
evacuated;  (2)  the  filament  is  burned  for  a 
shorr  period  at  about  2800°  Kelvin  to  clean 
the  surface  and  reduce  some  of  the  thoria 
within  the  filament  to  metallic  thorium;  (3) 


the  filament  is  burned  for  a longer  period  at 
about  2100°  Kelvin  to  form  a layer  of  thor- 
ium on  the  surface  of  the  tungsten;  (4)  the 
temperature  is  reduced  to  about  1600°  Kelvin 
and  some  pure  hydrocarbon  gas  is  admitted 
to  form  a layer  of  tungsten  carbide  on  the 
surface  of  the  tungsten.  This  layer  of  tungsten 
carbide  reduces  the  rate  of  thorium  evapora- 
tion from  the  surface  at  the  normal  operating 
temperature  of  the  filament  and  thus  increases 
the  operating  life  of  the  vacuum  tube.  Thor- 
ium evaporation  from  the  surface  is  a natural 
consequence  of  the  operation  of  the  thoriated- 
tungsten  filament.  The  carburized  layer  on  the 
tungsten  wire  plays  another  role  in  acting  as 
a reducing  agent  to  produce  new  thorium  from 
the  thoria  to  replace  that  lost  by  evapora- 
tion. This  new  thorium  continually  diffuses  to 
the  surface  during  the  normal  operation  of 
the  filament.  The  last  process,  (5),  in  the 
activation  of  a thoriated  tungsten  filament  con- 
sists of  re-evacuating  the  envelope  and  then 
burning  or  ageing  the  new  filament  for  a con- 
siderable period  of  time  at  the  normal  operat- 
ing temperature  of  approximately  1900°  K. 

One  thing  to  remember  about  any  type  of 
filament,  particularly  the  thoriated  type,  is 
that  the  emitter  deteriorates  practically  as 
fast  when  "standing  by”  (no  plate  current)  as 
it  does  with  any  normal  amount  of  emission 
load.  Also,  a thoriated  filament  may  be  either 
temporarily  or  permanently  damaged  by  a 
heavy  overload  which  may  srrip  the  surface 
layer  of  thorium  from  the  filament. 

Reactivating  Thoriated-tungsten  fila- 

Thorialed-Tungsten  ments  (and  only  thoriated- 
Filaments  tungsten  filaments)  which 

have  lost  emission  as 
a result  of  insufficient  filament  voltage,  a 
severe  temporary  overload,  a less  severe  ex- 
tended overload,  or  even  normal  operation 


HANDBOOK 


Types  of  Emitters  69 


Figure  2 

V-H-F  and  U-H-F  TUBE  TYPES 

The  tube  to  the  left  In  this  photograph  Is  a 955  *'ocorn**  triode.  The  6F4  acorn  trlode  Is  very  similar  In 
appearance  ta  the  955  hut  has  two  leads  brought  out  each  for  the  grid  and  for  the  plate  connection.  The 
second  tube  Is  a 446A  **lighthouse‘*  triode.  The  2C40,  2C43.  and  2C44  are  more  recent  examples  of  the 
some  type  tube  ond  are  essentially  the  same  in  external  appearance.  The  third  tube  from  the  left  is  a 
2C39  *‘oi‘/can"  tube.  This  tube  type  is  essentially  the  inverse  of  the  lighthouse  variety  since  the  cathode 
and  heater  connections  come  out  the  small  end  and  the  plate  Is  the  large  finned  radiator  on  the  large  end. 
The  use  of  the  finned  plate  radiator  makes  the  oilcan  tube  capable  of  approximately  10  times  as  much 
plate  dissipation  as  the  lighthouse  type.  The  tube  to  the  right  Is  the  4X150A  beam  tetrode.  This  tube,  a 
comparatively  recent  release,  Is  capable  of  somewhat  greater  power  output  than  any  of  the  other  tube 
types  shown,  and  Is  rated  for  full  output  at  500  Me.  and  at  reduced  output  at  frequencies  greater  than 

1000  Me. 


may  quite  frequently  be  reactivated  to  their 
original  characteristics  by  a process  similar 
to  that  of  the  original  activation.  However, 
only  filaments  which  have  not  approached  too 
close  to  the  end  of  their  useful  life  may  be 
successfully  reactivated. 

The  actual  process  of  reactivation  is  rel- 
atively simple.  The  tube  which  has  gone 
"flat”  is  placed  in  a socket  to  which  only  the 
two  filament  wires  have  been  connected.  The 
filament  is  then  "flashed”  for  about  20  to  40 
seconds  at  about  1%  times  normal  rated  volt- 
age. The  filament  will  become  extremely  bright 
during  this  time  and,  if  there  is  still  some 
thoria  left  in  the  tungsten  and  if  the  tube  did 
not  originally  fail  as  a result  of  an  air  leak, 
some  of  this  thoria  will  be  reduced  to  metallic 
thorium.  The  filament  is  then  burned  at  15  to 
25  pet  cent  overvoltage  for  from  30  minutes  to 
3 to  4 hours  to  bring  this  new  thorium  to  the 
surface. 

The  tube  should  then  be  tested  to  see  if  it 
shows  signs  of  renewed  life.  If  it  does,  but  is 
still  weak,  the  burning  process  should  be  con- 
tinued at  about  10  to  15  per  cent  overvoltage 
for  a few  more  hours.  This  should  bring  it 
back  almost  to  normal.  If  the  tube  checks  still 
very  low  after  the  first  attempt  at  reactivation, 
the  complete  process  can  be  repeated  as  a 
last  effort. 

The  Oxide-  The  most  efficient  of  all 

Coated  Filament  modern  filaments  is  the 
oxide-coated  type  which  con- 


sists of  a mixture  of  barium  and  strontium 
oxides  coated  upon  a nickel  alloy  wire  or 
strip.  This  type  of  filament  operates  at  a dull- 
red  to  orange-red  temperature  (1050°  to  1170° 
K)  at  which  temperature  it  will  emit  large 
quantities  of  electrons.  The  oxide-coated 
filament  is  somewhat  more  efficient  than  the 
thoriated-tungsten  type  in  small  sizes  and  it 
is  considerably  less  expensive  to  manufacture. 
For  this  reason  all  receiving  tubes  and  quite 
a number  of  the  low-powered  transmitting 
tubes  use  the  oxide-coated  filament.  Another 
advantage  of  the  oxide-coated  emitter  is  its 
extremely  long  life  — the  average  tube  can  be 
expected  to  tun  from  3000  to  5000  hours,  and 
when  loaded  very  lightly,  tubes  of  this  type 
have  been  known  to  give  50,000  houts  of  life 
before  their  characteristics  changed  to  any 
great  extent. 

Oxide  filaments  ate  unsatisfactory  for  use 
at  high  continuous  plate  voltages  because:  (1) 
their  activity  is  seriously  impaired  by  the 
high  temperature  necessary  to  de-gas  the  high- 
voltage  tubes  and,  (2)  the  positive  ion  bom- 
bardment which  takes  place  even  in  the  best 
evacuated  high-voltage  tube  causes  destruc- 
tion of  the  oxide  layer  on  the  surface  of  the 
filament. 

Oxide-coated  emitters  have  been  found  cap- 
able of  emitting  an  enormously  large  current 
pulse  with  a high  applied  voltage  for  a very 
short  period  of  time  without  damage.  This 
characteristic  has  proved  to  be  of  great  value 


70  Vacuum  Tube  Principles 


THE  RADIO 


Figure  3 

CUT-AWAY  DRAWING  OF  A 6C4  TRIODE 


in  radar  work.  For  example,  the  relatively 
small  cathode  in  a microwave  magnetron  may 
be  called  upon  to  deliver  25  to  50  amperes  at 
an  applied  voltage  of  perhaps  25,000  volts  for 
a period  in  the  order  of  one  microsecond. 
After  this  large  current  pulse  has  been  passed, 
plate  voltage  normally  will  be  removed  for 
1000  microseconds  or  more  so  that  the  cathode 
surface  may  be  restored  in  time  for  the  next 
pulse  of  current.  If  the  cathode  were  to  be 
subjected  to  a continuous  current  drain  of  this 
magnitude,  it  would  be  destroyed  in  an  ex- 
ceedingly short  period  of  time. 

The  activation  of  oxide-coated  filaments 
also  varies  with  tube  manufacturers  but  con- 
sists essentially  in  heating  the  wire  which  has 
been  coated  with  a mixture  of  barium  and 
strontium  carbonates  to  a temperature  of  about 
1500°  Kelvin  for  a time  and  then  applying  a 
potential  of  100  to  200  volts  through  a pro- 
tective resistor  to  limit  the  emission  current. 
This  process  reduces  the  carbonates  to  oxides 
thermally,  cleans  the  filament  surface  of 
foreign  materials,  and  activates  the  cathode 
surface. 

Reactivation  of  oxide-coated  filaments  is 
not  possible  since  there  is  always  more  than 
sufficient  reduction  of  the  oxides  and  diffusion 
of  the  metals  to  the  surface  of  the  filament  to 
meet  the  emission  needs  of  the  cathode. 

The  Heater  The  heater  type  cathode  was  de- 
Cathode  veloped  as  a result  of  the  re- 
quirement for  a type  of  emitter 
which  could  be  operated  from  alternating  cur- 
rent and  yet  would  not  introduce  a-c  ripple 
modulation  even  when  used  inlow-level  stages. 
It  consists  essentially  of  a small  nickel-alloy 
cylinder  with  a coating  of  strontium  and  bar- 
ium oxides  on  its  surface  similar  to  the  coat- 
ing used  on  the  oxide-coated  filament.  Inside 
the  cylinder  is  an  insulated  heater  element 
consisting  usually  of  a double  spiral  of  tung- 
sten wire.  The  heater  may  operate  on  any  volt- 


age from  2 to  117  volts,  although  6.3  is  the 

most  common  value.  The  heater  is  operated 
at  quite  a high  temperature  so  that  the  cathode 
itself  usually  may  be  brought  to  operating 
temperature  in  a matter  of  15  to  30  seconds. 
Heat-coupling  between  the  heater  and  the 
cathode  is  mainly  by  radiation,  although  there 
is  some  thermal  conduction  through  the  in- 
sulating coating  on  the  heater  wire,  as  this 
coating  is  also  in  contact  with  the  cathode 
thimble. 

Indirectly  heated  cathodes  are  employed  in 
all  a-c  operated  tubes  which  are  designed  to 
operate  at  a low  level  either  for  r-f  or  a-f  use. 
However,  some  receiver  power  tubes  use 
heater  cathodes  (6L6,  6V6,  6F6,  and  6K6-GT) 
as  do  some  of  the  low-power  transmitter  tubes 
(802,807,  815,  3E29,  2E26,  5763,  etc.).  Heater 
cathodes  ate  employed  almost  exclusively 
when  a number  of  tubes  ate  to  be  operated  in 
series  as  in  an  a.c.-d.c.  receiver.  A heater 
cathode  is  often  called  a uni-potential  cathode 
because  there  is  no  voltage  drop  along  its 
length  as  there  is  in  the  directly-heated  or 
filament  cathode. 

The  Bombardment  A special  bombardment  cath- 
Cothode  ode  is  employed  in  many  of 

the  new  high  powered  tele- 
vision transmitting  klystrons  (Eimac  3K  20,000 
LA).  The  cathode  takes  the  form  of  a tantalum 
diode,  heated  to  operating  temperature  by  the 
bombardment  of  electrons  from  a directly 
heated  filament.  The  cathode  operates  at  a 
positive  potential  of  2000  volts  with  respect 
to  the  filament,  and  a d-c  bombardment  cur- 
rent of  0.66  amperes  flows  between  filament 
and  cathode.  The  filament  is  designed  to 
operate  under  space-charge  limited  conditions. 
Cathode  temperature  is  varied  by  changing  the 
bombardment  potential  between  the  filament 
and  the  cathode. 

The  Emission  The  emission  of  electrons  from 
Equation  a heated  cathode  is  quite  sim- 

ilar to  the  evaporation  of  mole- 
cules from  the  surface  of  a liquid.  The  mole- 
cules which  leave  the  surface  are  those  having 
sufficient  kinetic  (heat)  energy  to  overcome 
the  forces  at  the  surface  of  the  liquid.  As  the 
temperature  of  the  liquid  is  raised,  the  aver- 
age velocity  of  the  molecules  is  increased, 
and  a greater  number  of  molecules  will  acquire 
sufficient  energy  to  be  evaporated.  The  evapor- 
ation of  electrons  from  the  surface  of  a rher- 
mionic  emitter  is  similarly  a function  of  aver- 
age electron  velocity,  and  hence  is  a function 
of  the  temperature  of  the  emitter. 

Electron  emission  per  unit  area  of  emitting 
surface  is  a function  of  the  temperature  T 
in  degrees  Kelvin,  the  work  function  of  the 
emitting  surface  b (which  is  a measure  of  the 


HANDBOOK 


Thermionic  Emission  71 


CUT-AWAY  DRAWING  OF  A 6CB«  PENTODE 


TYPE  6W4-GT 

Ef  = 6.3  VOLTS 

1 

4 

V 

/ 

/ 

A 

/ 

y 

0 10  20  30  40  50 


D.C.  PLATE  VOLTS 


surface  forces  of  the  material  and  hence  of 
the  energy  required  of  the  electron  before  it 
may  escape),  and  of  the  constant  A which 
also  varies  with  the  emitting  surface.  The  re- 
lationship  between  emission  current  in  am- 
peres per  square  centimeter,  /,  and  the  above 
quantities  can  be  expressed  as: 

/ = 

Secondary  The  bombarding  of  most  metals 
Emission  and  a few  insulators  by  electrons 
will  result  in  the  emission  of  other 
electrons  by  a process  called  secondary  emis- 
sion. The  secondary  electrons  are  literally 
knocked  f rom  the  surface  layers  of  the  bom- 
barded material  by  the  primary  electrons  which 
strike  the  material.  The  number  of  secondary 
electrons  emitted  per  primary  electron  varies 
from  a very  small  percentage  to  as  high  as 
5 to  10  secondary  electrons  per  primary. 

The  phenomena  of  secondary  emission  is 
undesirable  for  most  thermionic  electron  tubes. 
However,  the  process  is  used  to  advantage  in 
certain  types  of  electron  tubes  such  as  the 
image  orthicon  (TV  camera  tube)  and  the 
electron-multiplier  type  of  photo-electric  cell. 
In  types  of  electron  tubes  which  make  use  of 
secondary  emission,  such  as  the  type  931 
photo  cell,  the  secondary-electron-emitting 
surfaces  are  specially  treated  to  provide  a 
high  ratio  of  secondary  to  primary  electrons. 
Thus  a high  degree  of  current  amplification  in 
the  electron-multiplier  section  of  the  tube  is 
obtained. 

The  Space  As  a cathode  is  heated  so  that 
Charge  Effect  it  begins  to  emit,  those  elec- 
trons which  have  been  dis- 
charged into  the  surrounding  space  form  a 
negatively  charged  cloud  in  the  immediate 
vicinity  of  the  cathode.  This  cloud  of  electrons 
around  the  cathode  is  called  the  space  charge. 
The  electrons  comprising  the  charge  are  con- 
tinuously changing,  since  those  electrons 
making  up  the  original  charge  fall  back  into 


Figure  5 

AVERAGE  PLATE  CHARACTERISTICS 
OF  A POWER  DIODE 


the  cathode  and  are  replaced  by  others  emitted 
by  it. 

4-2  The  Diode 

If  a cathode  capable  of  being  heated  either 
indirectly  or  directly  is  placed  in  an  evacuated 
envelope  along  with  a plate,  such  a two- 
element  vacuum  tube  is  called  a diode.  The 
diode  is  the  simplest  of  all  vacuum  tubes  and 
is  the  fundamental  type  from  which  all  the 
others  are  derived. 

Characteristics  When  the  cathode  within  a 
of  the  Diode  diode  is  heated,  it  will  be 
found  that  a few  of  the  elec- 
trons leaving  the  cathode  will  leave  with  suf- 
ficient velocity  to  reach  the  plate.  If  the  plate 
is  electrically  connected  back  to  the  cathode, 
the  electrons  which  have  had  sufficient  veloc- 
ity to  arrive  at  the  plate  will  flow  back  to  the 
cathode  through  the  external  circuit.  This 
small  amount  of  initial  plate  current  is  an 
effect  found  in  all  two-element  vacuum  tubes. 

If  a battery  or  other  source  of  d-c  voltage 
is  placed  in  the  external  circuit  between  the 
plate  and  cafhode  so  that  it  places  a positive 
potential  on  the  plate,  the  flow  of  current  from 
the  cathode  to  plate  will  be  increased.  This  is 
due  to  the  strong  attraction  offered  by  theposi- 
tively  charged  plate  for  any  negatively  charged 
particles  (figure  5). 

Space-Charge  Limited  At  moderate  values  of 
Current  plate  voltage  the  cur- 

rent flow  from  cathode 
to  anode  is  limited  by  the  space  charge  of 
electrons  around  the  cathode.  Increased  values 


72  Vacuum  Tube  Principles 


THE  RADIO 


Figure  6 

MAXIMUM  SPACE-CHARGE-LIMITED 
EMISSION  FOR  DIFFERENT 
TYPES  OF  EMITTERS 


of  plate  voltage  will  tend  to  neutralize  a 
greater  portion  of  the  cathode  space  charge 
and  hence  will  cause  a greater  current  to  flow. 

Under  these  conditions,  with  plate  current 
limited  by  the  cathode  space  charge,  the  plate 
current  is  not  linear  with  plate  voltage.  In 
fact  it  may  be  stated  in  general  that  the  plate- 
current  flow  in  electron  tubes  does  not  obey 
Ohm’s  Law.  Rather,  plate  current  increases  as 
the  three-halves  power  of  the  plate  voltage. 
The  relationship  between  plate  voltage,  E, 
and  plate  current,  f,  can  be  expressed  as: 

I = K EV^ 

where  K is  a constant  determined  by  the 
geometry  of  the  element  structure  within  the 
electron  tube. 

Plate  Current  As  plate  voltage  is  raised  to 
Saturation  the  potential  where  the  cath- 
ode space  charge  is  neutral- 
ized, all  the  electrons  that  the  cathode  is  cap- 
able of  emitting  are  being  attracted  to  the 
plate.  The  electron  tube  is  said  then  to  have 
reached  saturation  plate  current.  Further  in- 
crease in  plate  voltage  will  cause  only  a 
relatively  small  increase  in  plate  current.  The 
initial  point  of  plate  current  saturation  is 
sometimes  called  the  point  of  Maximum  Space- 
Charge-Limited  Emission  (MSCLE). 

The  degree  of  flattening  in  the  plate-voltage 
plate-current  curve  after  the  MSCLE  point  will 
vary  with  different  types  of  cathodes.  This  ef- 
fect is  shown  in  figure  6.  The  flattening  is 
quite  sharp  with  a pure  tungsten  emitter.  With 
thotiated  tungsten  the  flattening  is  smoothed 
somewhat,  while  with  an  oxide-coated  cathode 
the  flattening  is  quite  gradual.  The  gradual 
saturation  in  emission  with  an  oxide-coated 
emitter  is  generally  considered  to  result  from 


@ (D  © 


Figure  7 

ACTION  OF  THE  GRID  IN  A TRIODE 

(A)  shows  the  frtode  tube  with  cutoff  bias  on 
the  grid.  Note  that  all  the  electrons  emitted 
by  the  cathode  remain  inside  the  grid  mesh. 

(B)  shows  the  same  tube  with  an  /ntermec/iofe 
value  of  bias  on  the  grid.  Note  the  medium 
value  of  plate  current  and  the  fact  that  there 
is  a reserve  of  electrons  remaining  within  the 
grid  mesh.  (C)  shows  the  operation  with  a 
relatively  small  amount  of  bias  which  with 
certain  tube  types  will  allow  substantially  all 
the  electrons  emitted  by  the  cathode  to  reach 
the  plate.  Emission  is  said  to  be  saturated  in 
this  case.  In  a majority  of  tube  types  a high 
value  of  positive  grid  voltage  is  required  be- 
fore pi  ate~current  sofurotion  takes  place. 


a lowering  of  the  surface  work  function  by  the 
field  at  the  cathode  resulting  from  the  plate 
potential. 

Electron  Energy  The  current  flowing  in  the 
Dissipation  plate-cathode  space  of  a con- 

ducting electron  tube  repre- 
sents the  energy  required  to  accelerate  elec- 
trons from  the  zero  potential  of  the  cathode 
space  charge  to  the  potential  of  the  anode. 
Then,  when  these  accelerated  electrons  strike 
the  anode,  the  energy  associated  with  their 
velocity  is  immediately  released  to  the  anode 
structure.  In  normal  electron  tubes  this  energy 
release  appears  as  heating  of  the  plate  or 
anode  structure. 


4-3  The  Triode 

If  an  element  consisting  of  a mesh  or  spiral 
of  wire  is  inserted  concentric  with  the  plate 
and  between  the  plate  and  the  cathode,  such 
an  element  will  be  able  to  control  by  electro- 
static action  the  cathode-to-plate  current  of 
the  tube.  The  new  element  is  called  a grid,  and 
a vacuum  tube  containing  a cathode,  grid,  and 
plate  is  commonly  called  a triode. 

Action  of  If  this  new  element  through  which 
the  Grid  the  electrons  must  pass  in  their 
course  from  cathode  to  plate  is  made 
negative  with  respect  to  the  cathode,  the  nega- 


HANDBOOK 


Triode  Characteristics  73 


Figure  8 

NEGATIVE-GRID  CHARACTERISTICS  (Ip 
VS.  Ep  CURVES)  OF  A TYPICAL 
TRIODE 

Average  plate  characteristics  of  this  type 
are  most  commonly  used  in  determining  the 
Class  A operating  characteristics  of  a 
triode  amplifier  stage. 


five  charge  on  this  grid  will  effectively  repel 
the  negatively  charged  electrons  (like  charges 
repel;  unlike  charges  attract)  back  into  the 
space  charge  surrounding  the  cathode.  Hence, 
the  number  of  electrons  which  are  able  to  pass 
through  the  grid  mesh  and  reach  the  plate  will 
be  reduced,  and  the  plate  current  will  be  re- 
duced accordingly.  If  the  charge  on  the  grid 
is  made  sufficiently  negative,  all  the  electrons 
leaving  the  cathode  will  be  repelled  back  to 
it  and  the  plate  current  will  be  reduced  to  zero. 
Any  d-c  voltage  placed  upon  a grid  is  called 
a bias  (especially  so  when  speaking  of  a con- 
trol grid).  The  smallest  negative  voltage  which 
will  cause  cutoff  of  plate  current  at  a particu- 
lar plate  voltage  is  called  the  value  of  cutoff 
bias  (figure  7). 

Amplification  The  amount  of  plate  current  in  a 
Factor  triode  is  a result  of  the  net  field 

at  the  cathode  from  interaction 
between  the  field  caused  by  the  grid  bias  and 
that  caused  by  the  plate  voltage.  Hence,  both 
grid  bias  and  plate  voltage  affect  the  plate 
current.  In  all  normal  tubes  a small  change  in 
grid  bias  has  a considerably  greater  effect 
than  a similar  change  in  plate  voltage.  The 
ratio  between  the  change  in  grid  bias  and  the 
change  in  plate  current  which  will  cause  the 
same  small  change  in  plate  current  is  called 
the  amplification  factor  or  p of  the  electron 
tube.  Expressed  as  an  equation: 

AEp 


with  ip  constant  (A  represents  a small  incre- 
ment). 

The  n can  be  determined  experimentally  by 
making  a small  change  in  grid  bias,  thus 
slightly  changing  the  plate  current.  The  plate 
current  is  then  returned  to  the  original  value 
by  making  a change  in  the  plate  voltage.  The 
ratio  of  the  change  in  plate  voltage  to  the 
change  in  grid  voltage  is  the  p of  the  tube 
under  the  operating  conditions  chosen  for  the 
test. 

Current  Flow  In  a diode  it  was  shown  that 
in  a Triode  the  electrostatic  field  at  the 
cathode  was  proportional  to 
the  plate  potential,  Ep,  and  that  the  total 
cathode  current  was  proportional  to  the  three- 
halves  power  of  the  plate  voltage.  Similiarly, 
in  a triode  it  can  be  shown  that  the  field  at 
the  cathode  space  charge  is  proportional  to 
the  equivalent  voltage  (Eg  + Ep//x),  where 
the  amplification  factor,  fi,  actually  represents 
the  relative  effectiveness  of  grid  potential  and 
plate  potential  in  producing  a field  at  the 
cathode. 

It  would  then  be  expected  that  the  cathode 
current  in  a triode  would  be  proportional  to 
the  three-halves  power  of  (Eg  +Ep/fi).  The 
cathode  current  of  a triode  can  be  represented 
with  fair  accuracy  by  the  expression: 

/ Ep 

Cathode  current  = K ^Eg  -f j 

where  K is  a constant  determined  by  element 
geometry  within  the  triode. 

Plate  Resistance  The  plate  resistance  of  a 
vacuum  tube  is  the  ratio  of  a 
change  in  plate  voltage  to  the  change  in  plate 
current  which  the  change  in  plate  voltage 
produces.  To  be  accurate,  the  changes  should 
be  very  small  with  respect  to  the  operating 
values.  Expressed  as  an  equation: 

AEp 

R„  = E.  = constant,  A = small 

Al  * • 

cUp  increment 

The  plate  resistance  can  also  be  determined 
by  tbe  experiment  mentioned  above.  By  noting 
the  change  in  plate  current  as  it  occurs  when 
the  plate  voltage  is  changed  (grid  voltage 
held  constant),  and  by  dividing  the  latter  by 
the  former,  the  plate  resistance  can  be  deter- 
mined. Plate  resistance  is  expressed  in  Ohms. 

Transconduetance  The  mutual  conductance, 
also  referred  to  as  trans- 
conductance,  is  the  ratio  of  a change  in  the 
plate  current  to  the  change  in  grid  voltage 
which  brought  about  the  plate  current  change, 
the  plate  voltage  being  held  constant.  Ex- 
pressed as  an  equation: 


74  Vacuum  Tube  Principles 


THE  RADIO 


Figure  9 

POSITIVE-GRID  CHARACTERISTICS 
(Ip  VS.  Eg)  OF  A TYPICAL  TRIODE 

P/ofe  characteristics  of  this  type  are  most 
commonly  used  in  determining  the  pulse-signal 
operating  characteristics  of  a triode  amplifier 
stage.  Note  the  large  emission  capability  of 
the  oxide-coated  heater  cathode  in  tubes  of  the 
general  type  of  the  6JS. 


Go, 


AEg 


Ep  = constant,  A = small 

increment 


The  transconductance  is  also  numerically 
equal  to  the  amplification  factor  divided  by 
rhe  plate  resistance.  Gn,  = (i/Rp. 


Transconductance  is  most  commonly  ex- 
pressed in  microreciprocal-ohms  or  micro- 
mhos. However,  since  transconductance  ex- 
presses change  in  plate  current  as  a function 
of  a change  in  grid  voltage,  a tube  is  often 
said  to  have  a transconductance  of  so  many 
milliamperes-per-volt.  If  the  transconductance 
in  milliamperes-per-volt  is  multiplied  by  1000 
it  will  then  be  expressed  in  micromhos.  Thus 
the  transconductance  of  a 6A3  could  be  called 
either  5-25  ma./volt  or  5250  micromhos. 


Characteristic  Curves  The  operating  character- 
of  a Triode  Tube  istics  of  a triode  tube 

may  be  summarized  in 
three  sets  of  curves:  The  Ip  vs.  Ep  curve 
(figure  8),  the  Ip  vs.  Eg  curve  (figure  9)  and 
the  Ep  VS.  Eg  curve  (figure  10).  The  plate 
resistance  (Rp)  of  the  tube  may  be  observed 
from  the  Ip  vs.  Ep  curve,  the  transconductance 
(Gm)  may  be  observed  from  the  Ip  vs.  Eg  curve, 
and  the  amplification  factor  (p)  may  be  deter- 
mined from  the  Ep  vs.  Eg  curve. 

The  Load  Line  A load  line  is  a graphical 
representation  of  the  voltage 
on  the  plate  of  a vacuum  tube,  and  the  current 


Figure  10 

CONSTANT  CURRENT  (Ep  VS.  Eg) 
CHARACTERISTICS  OF  A 
TYPICAL  TRIODE  TUBE 

This  type  of  graphical  representation  Is  used 
for  Class  C amplifier  calculations  since  the 
operating  characteristic  of  a Class  C amplifier 
ie  a straight  line  when  drawn  upon  a constant 
current-  graph. 


passing  through  the  plate  circuit  of  the  tube 
for  various  values  of  plate-load  resistance  and 
plate-supply  voltage.  Figure  11  illustrates  a 
triode  tube  with  a resistive  plate  load,  and  a 
supply  voltage  of  300  volts.  The  voltage  at 
the  plate  of  the  tube  (ep)  may  be  expressed 
as: 

= Ep  “(ip  X El] 

where  Ep  is  the  plate  supply  voltage,  ip  is  the 
plate  current,  and  Rl  is  the  load  resistance  in 
ohms. 

Assuming  various  values  of  ip  flowing  in 
the  circuit,  controlled  by  the  internal  resist- 
ance of  the  tube,  (a  function  of  the  grid  bias) 
values  of  plate  voltage  may  be  plotted  as 
shown  for  each  value  of  plate  current  (ip).  The 
line  connecting  these  points  is  called  the 
load  line  for  the  particular  value  of  plate-load 
resistance  used.  The  slope  of  the  load  line  is 
equal  to  the  ratio  of  the  lengths  of  the  vertical 
and  horizontal  projections  of  any  segment  of 
the  load  line.  For  this  example  it  is: 

.01  - .02  1 

Slope  —.0001  

100  - 200  10,000 

The  slope  of  the  load  line  is  equal  to 
— 1/Rl.  At  point  A on  the  load  line,  the  volt- 
age across  the  tube  is  zero.  This  would  be 
true  for  a perfect  tube  with  zero  internal  volt- 
age drop,  or  if  the  tube  is  short-circuited  from 
cathode  to  plate.  Point  B on  the  load  line 
corresponds  to  the  cutoff  point  of  the  tube, 
where  no  plate  current  is  flowing.  The  op- 
erating range  of  the  tube  lies  between  these 
two  extremes.  For  additional  information  re- 


HANDBOOK 


Triode  Load  Line  75 


0 

300 

5 

250 

10 

200 

IS 

1 50 

20 

100 

25 

50 

20 

0 

Figure  1 1 

The  static  load  lino  for  a typical  triads 
tubs  with  a plots  load  rssistancs  of  70,000 
ohms. 


garding  dynamic  load  lines,  the  reader  is 
referred  to  the  Radiotron  Designer’s  Handbook, 
4th  edition,  distributed  by  Radio  Corporation 
of  America. 


Application  of  Tube  As  an  example  of  the  ap- 
Characteristics  plication  of  tube  charac- 

teristics, the  constants 
of  the  triode  amplifier  circuit  shown  in  figure 
12  may  be  considered.  The  plate  supply  is 


Figure  12 

TRIODE  TUBE  CONNECTED  FOR  DETER- 
MINATION OF  PLATE  CIRCUIT  LOAD 
LINE,  AND  OPERATING  PARAMETERS 
OF  THE  CIRCUIT 


300  volts,  and  the  plate  load  is  8,000  ohms. 
If  the  tube  is  considered  to  be  an  open  circuit 
no  plate  current  will  flow,  and  there  is  no 
voltage  drop  across  the  plate  load  resistor, 
Rl-  The  plate  voltage  on  the  tube  is  therefore 
300  volts.  If,  on  the  other  hand,  the  tube  is 
considered  to  be  a short  circuit,  maximum 
possible  plate  current  flows  and  the  full  300 
volt  drop  appears  across  Rj,.  The  plate  volt- 
age is  zero,  and  the  plate  current  is  300/8,000, 
or  37.5  milliamperes.  These  two  extreme  con- 
ditions define  the  load  line  on  the  Ip  vs.  Ep 
characteristic  curve,  figure  13. 

For  this  application  the  grid  of  the  tube  is 
returned  to  a steady  biasing  voltage  of  -4 
volts.  The  steady  or  quiescent  operation  of  the 
tube  is  determined  by  the  intersection  of  the 
load  line  with  the  -4  volt  curve  at  point  Q. 
By  projection  from  point  Q through  the  plate 


Figure  13 

APPLICATION  OF  Ip  VS.  Ep 
CHARACTERISTICS  OF 
VACUUM  TUBE 


76  Vacuum  Tube  Principles 


THE  RADIO 


Figure  14 

POLARITY  REVERSAL  BETWEEN  GRID 
AND  PLATE  VOLTAGES 


current  axis  it  is  found  that  the  value  of  plate 
current  with  no  signal  applied  to  the  grid  is 
12.75  milliamperes.  By  projection  from  point 
Q through  the  plate  voltage  axis  it  is  found 
that  the  quiescent  plate  vpltage  is  198  volts. 
This  leaves  a drop  of  102  volts  across  Rl 
which  is  borne  out  by  the  relation  0.01275  x 
8,000  = 102  volts. 

An  alternating  voltage  of  4 volts  maximum 
swing  about  the  normal  bias  value  of  -4  volts 
is  applied  now  to  the  grid  of  the  ttiode  ampli- 
fier. This  signal  swings  the  grid  in  a positive 
direction  to  0 volts,  and  in  a negative  direction 
to  -8  volts,  and  establishes  the  operating 
region  of  the  tube  along  the  load  line  between 
points  A and  B.  Thus  the  maxima  and  minima 
of  the  plate  voltage  and  plate  current  are 
established.  By  projection  from  points  A and 
B through  the  plate  current  axis  the  maximum 
instantaneous  plate  current  is  found  to  be 
18.25  milliamperes  and  the  minimum  is  7.5 
milliamperes.  By  projections  from  points  A and 
B through  the  plate  voltage  axis  the  minimum 
instantaneous  plate  voltage  swing  is  found  to 
be  154  volts  and  the  maximum  is  240  volts. 

By  this  graphical  application  of  the  Ip  vs. 
Ep  characteristic  of  the  6SN7  ttiode  the  opera- 
tion of  the  circuit  illustrated  in  figure  12  be- 


Figure  15 

SCHEMATIC  REPRESENTATION 
OF  INTERELECTRODE 
CAPACITANCE 


comes  apparent.  A voltage  variation  of  8 volts 
(peak-to-peak)  on  the  grid  produces  a variation 
of  84  volts  at  the  plate. 

Polarity  Inversion  When  the  signal  voltage  ap- 
plied to  the  grid  has  its 
maximum  positive  instantaneous  value  the 
plate  current  is  also  maximum.  Reference  to 
figure  12  shows  that  this  maximum  plate  cur- 
rent flows  through  the  plate  load  resistor  R^, 
producing  a maximum  voltage  drop  across  it. 
The  lower  end  of  Rl  is  connected  to  the  plate 
supply,  and  is  therefore  held  at  a constant 
potential  of  300  volts.  With  maximum  voltage 
drop  across  the  load  resistor,  the  upper  end  of 
Rl  is  at  a minimum  instantaneous  voltage. 
The  plate  of  the  tube  is  connected  to  this  end 
of  Rl  and  is  therefore  at  the  same  minimum 
instantaneous  potential. 

This  polarity  reversal  between  instantaneous 
grid  and  plate  voltages  is  further  clarified  by 
a consideration  of  Kirchhoff’s  law  as  it  ap- 
plies to  series  resistance.  The  sum  of  the  IR 
drops  around  the  plate  circuit  must  at  all  times 
equal  the  supply  voltage  of  300  volts.  Thus 
when  the  instantaneous  voltage  drop  across 
Rl  is  maximum,  the  voltage  drop  across  the 
tube  is  minimum,  and  their  sum  must  equal 
300  volts.  The  variations  of  grid  voltage,plate 
current  and  plate  voltage  about  their  steady 
state  values  is  illustrated  in  figure  14. 

Interelectrode  Capacitance  always  exists  be- 
Copocitonce  tween  any  two  pieces  of  metal 
separated  by  a dielectric.  The 
exact  amount  of  capacitance  depends  upon  the 
size  of  the  metal  pieces,  the  dielectric  be- 
tween them,  and  the  type  of  dielectric.  The 
electrodes  of  a vacuum  tube  have  a similar 
characteristic  known  as  the  interelectrode 
capacitance,  illustrated  in  figure  15.  These 
direct  capacities  in  a triode  are;  grid-to- 
cathode  capacitance,  grid-to-plate  capacitance, 
and  plate-to-cathode  capacitance.  The  inter- 
electrode  capacitance,  though  very  small,  has 
a coupling  effect,  and  often  can  cause  un- 
balance in  a particular  circuit.  At  very  high 


HANDBOOK 


Tetrodes  and  Pentodes  77 


Figure  16 

TYPICAL  Ip  VS.  Ep  TETRODE 
CHARACTERISTIC  CURVES 


Figure  17 

TYPICAL  Ip  VS.  Ep  PENTODE 
CHARACTERISTIC  CURVES 


frequencies  (v-h-f),  interelectrode  capacities 
become  very  objectionable  and  prevent  the  use 
of  conventional  tubes  at  these  frequencies. 
Special  v-h-f  tubes  must  be  used  which  are 
characterized  by  very  small  electrodes  and 
close  internal  spacing  of  the  elements  of  the 
tube. 

4-4  Tetrode  or  Screen  Grid  Tubes 

Many  desirable  characteristics  can  be  ob- 
tained in  a vacuum  tube  by  the  use  of  more 
than  one  grid.  The  most  common  multi-element 
tube  is  the  tetrode  (four  electrodes).  Other 
tubes  containing  as  many  as  eight  electrodes 
are  available  for  special  applications. 

The  Tetrode  The  quest  for  a simple  and  easily 
usable  method  of  eliminating  the 
effects  of  the  grid-to-plate  capacitance  of  the 
triode  led  to  the  development  of  the  screen- 
grid  tube  or  tetrode.  Wben  another  grid  is 
added  between  the  grid  and  plate  of  a vacuum 
tube  the  tube  is  called  a tetrode,  and  because 
the  new  grid  is  called  a screen,  as  a result  of 
its  screening  or  shielding  action,  the  tube  is 
often  called  a screen-grid  tube.  The  inter- 
posed screen  grid  acts  as  an  electrostatic 
shield  between  the  grid  and  plate,  with  the 
consequence  that  the  grid-to-plate  capacitance 
is  reduced.  Although  the  screen  grid  is  main- 
tained at  a positive  voltage  with  respect  to 
the  cathode  of  the  tube,  it  is  maintained  at 
ground  potential  with  respect  to  t.f.  by  means 
of  a by-pass  capacitor  of  very  low  reactance 
at  the  frequency  of  operation. 

In  addition  to  the  shielding  effect,  the 
screen  grid  serves  another  very  useful  purpose. 
Since  the  screen  is  maintained  at  a positive 
potential,  it  serves  to  increase  or  accelerate 
the  flow  of  electrons  to  the  plate.  There  being 
large  openings  in  the  screen  mesh,  most  of 


the  electrons  pass  through  it  and  on  to  the 
plate.  Due  also  to  the  screen,  the  plate  cur- 
rent is  largely  independent  of  plate  voltage, 
thus  making  for  high  amplification.  When  the 
screen  voltage  is  held  at  a constant  value,  it 
is  possible  to  make  large  changes  in  plate 
voltage  without  appreciably  affecting  the  plate 
current,  (figure  16). 

When  the  electrons  from  the  cathode  ap- 
proach the  plate  with  sufficient  velocity,  they 
dislodge  electrons  upon  striking  the  plate. 
This  effect  of  bombarding  the  plate  with  high 
velocity  electrons,  with  the  consequent  dis- 
lodgement  of  other  electrons  from  the  plate, 
gives  rise  to  the  condition  of  secondary  emis- 
sion which  has  been  discussed  in  a previous 
paragraph.  This  effect  can  cause  no  particular 
difficulty  in  a triode  because  the  secondary 
electrons  so  emitted  are  eventually  attracted 
back  to  the  plate.  In  the  screen-grid  tube,  how- 
ever, the  screen  is  close  to  the  plate  and  is 
maintained  at  a positive  potential.  Thus,  the 
screen  will  attract  these  electrons  which  have 
been  knocked  from  the  plate,  particularly  when 
the  plate  voltage  falls  to  a lower  value  than 
the  screen  voltage,  with  the  result  that  the 
plate  current  is  lowered  and  the  amplification 
is  decreased. 

In  the  application  of  tetrodes,  it  is  neces- 
sary to  operate  the  plate  at  a high  voltage  in 
relation  to  the  screen  in  order  to  overcome 
these  effects  of  secondary  emission. 

The  Pentode  The  undesirable  effects  of  sec- 
ondary emission  from  the  plate 
can  be  greatly  reduced  if  yet  another  element 
is  added  between  the  screen  and  plate.  This 
additional  element  Is  called  a suppressor,  and 
tubes  in  which  it  is  used  are  called  pentodes. 
The  suppressor  grid  is  sometimes  connected 

to  the  cathode  within  the  tube;  sometimes  it  is 
brought  out  to  a connecting  pin  on  the  tube 
base,  but  in  any  case  it  is  established  nega- 


78  Vacuum  Tube  Principles 


THE  RADIO 


Figure  18 

REMOTE  CUTOFF  GRID  STRUCTURE 


Figure  19 

ACTION  OF  A REMOTE  CUTOFF 
GRID  STRUCTURE 


tive  with  respect  to  the  minimum  plate  volt- 
age. The  secondary  electrons  that  would  travel 
to  the  screen  if  there  were  no  suppressor  are 
diverted  back  to  the  plate.  The  plate  current 
is,  therefore,  not  reduced  and  the  amplifica- 
tion possibilities  are  increased  (figure  17). 

Pentodes  for  audio  applications  are  de- 
signed so  that  the  suppressor  increases  the 
limits  to  which  the  plate  voltage  may  swing; 
therefore  the  consequent  power  output  and 
gain  can  be  very  great.  Pentodes  for  radio- 
frequency service  function  in  such  a manner 
that  the  suppressor  allows  high  voltage  gain, 
at  the  same  time  permitting  fairly  high  gain 
at  low  plate  voltage.  This  holds  true  even  if 
the  plate  voltage  is  the  seime  or  slightly  lower 
than  the  screen  voltage. 

Remote  Cutoff  Remote  cutojf  tubes  (variable 
Tubes  mu)  are  screen  grid  tubes  in 

which  the  control  grid  struc- 
ture has  been  physically  modified  so  as  to 
cause  the  plate  current  of  the  tube  to  drop  off 
gradually,  rather  than  to  have  a well  defined 
cutoff  point  (figure  18).  A non-uniform  control 
grid  structure  is  used,  so  that  the  amplifica- 
tion factor  is  different  for  different  parts  of  the 
control  grid. 

Remote  cutoff  tubes  are  used  in  circuits 
where  it  is  desired  to  control  the  amplification 
by  varying  the  control  grid  bias.  The  charac- 
teristic curve  of  an  ordinary  screen  grid  tube 
has  considerable  curvature  neat  the  plate  cur- 
rent cutoff  point,  while  the  curve  of  a remote 
cutoff  tube  is  much  mote  linear  (figure  19). 
The  remote  cutoff  tube  minimizes  cross- 
talk interference  that  would  otherwise  be 
produced.  Examples  of  remote  cutoff  tubes 
ate:  6BD6,  6K7,  6SG7  and  6SK7. 

Beam  Power  A beam  power  tube  makes  use 
Tubes  of  another  method  for  suppressing 

secondary  emission.  In  this  tube 
there  ate  four  electrodes;  a cathode,  a grid,  a 
screen,  and  a plate,  so  spaced  and  placed  that 
secondary  emission  from  the  plate  is  sup- 
pressed without  actual  power  loss.  Because 


of  the  manner  in  which  the  electrodes  are 
spaced,  the  electrons  which  travel  to  the 
plate  ate  slowed  down  tdten  the  plate  voltage 
is  low,  almost  to  zero  velocity  in  a certain 
region  between  screen  and  plate.  For  this 
reason  the  electrons  form  a stationary  cloud, 
or  space  charge.  The  effect  of  this  space 
charge  is  to  repel  secondary  electrons  emitted 
from  the  plate  and  thus  cause  them  to  return 
to  the  plate.  In  this  way,  secondary  emission 
is  suppressed. 

Another  feature  of  the  beam  power  tube  is 
the  low  current  drawn  by  the  screen.  The 
screen  and  the  grid  are  spiral  wires  wound  so 
that  each  turn  in  the  screen  is  shaded  from 
the  cathode  by  a grid  turn.  This  alignment  of 
the  screen  and  the  grid  causes  the  electrons 
to  travel  in  sheets  between  the  turns  of  the 
screen  so  that  very  few  of  them  strike  the 
screen  itself.  This  formation  of  the  electron 
stream  into  sheets  or  beams  increases  the 
charge  density  in  the  screen-plate  region  and 
assists  in  the  creation  of  the  space  charge  in 
this  region. 

Because  of  the  effective  suppressor  action 
provided  by  the  space  charge,  and  because  of 
the  low  current  drawn  by  the  screen,  the  beam 
power  tube  has  the  advantages  of  high  power 
output,  high  power-sensitivity,  and  high  ef- 
ficiency. The  6L6  is  such  a beam  power  tube, 
designed  for  use  in  the  power  amplifier  stages 
of  receivers  and  speech  amplifiers  or  modulat- 
ors. Larger  tubes  employing  the  beam-power 
principle  ate  being  made  by  various  manu- 
facturers for  use  in  the  radio-frequency  stages 
of  transmitters.  These  tubes  feature  extremely 
high  power- sensitivity  (a  very  small  amount 
of  driving  power  is  required  for  a large  out- 
put), good  plate  efficiency,  and  low  grid-to- 
plate  capacitance.  Examples  of  these  tubes 
are  813,  4-250A,  4X150A,  etc. 

Grid-Screen  The  grid- screen  mu  factor  (psg) 
Mu  Factor  is  analogous  to  the  amplification 
factor  in  a triode,  except  that 
the  screen  of  a pentode  or  tetrode  is  sub- 


HANDBOOK 


Mixer  and  Converter  Tubes  79 


stituted  for  the  plate  of  a triode.  denotes 
the  ratio  of  a change  in  grid  voltage  to  a 
change  in  screen  voltage,  each  of  which  will 
produce  the  same  change  in  screen  current. 
Expressed  as  an  equation: 

AEsg 

Psg  = "T ^sg  = constant,  A = small 

increment 

The  grid- screen  mu  factor  is  important  in 
determining  the  operating  bias  of  a tetrode 
or  pentode  tube.  The  relationship  between  con- 
trol-grid potential  and  screen  potential  deter- 
mines the  plate  current  of  the  tube  as  well  as 
the  screen  current  since  the  plate  current  is 
essentially  independent  of  the  plate  voltage 
in  tubes  of  this  type.  In  other  words,  when 
the  tube  is  operated  at  cutoff  bias  as  deter- 
mined by  the  screen  voltage  and  the  grid- 
screen  mu  factor  (determined  in  the  same  way 
as  with  a triode,  by  dividing  the  operating 
voltage  by  the  mu  factor)  the  plate  current 
will  be  substantially  at  cutoff,  as  will  be  the 
screen  current.  The  grid-screen  mu  factor  is 
numerically  equal  to  the  amplification  factor 
of  the  same  tetrode  or  pentode  tube  when 
it  is  triode  connected. 

Current  Flow  The  following  equation  is  the 
In  Tetrodes  expression  fortoti  cathode  cur- 
ond  Pentodes  rent  in  a triode  tube.  The  ex- 
pression for  the  total  cathode 
current  of  a tetrode  and  a pentode  tube  is  the 
same,  except  that  the  screen-grid  voltage  and 
the  grid-screen  /t-factor  are  used  in  place  of 
the  plate  voltage  and  p of  the  triode. 

/ E 

Cathode  current  = K ( Eg  ^ 1 
' Fsg  7 

Cathode  current,  of  course,  is  the  sum  of  the 
screen  and  plate  current,  plus  control  grid  cur- 
rent in  the  event  that  the  control  grid  is  posi- 
tive with  respect  to  the  cathode.  It  will  be 
noted  that  total  cathode  current  is  independent 
of  plate  voltage  in  a tetrode  or  pentode.  Also, 
in  the  usual  tetrode  or  pentode  the  plate  cur- 
rent is  substantially  independent  of  plate 
voltage  over  the  usual  operating  range — which 
means  simply  that  the  effective  plate  resist- 
ance of  such  tubes  is  relatively  high.  How- 
ever, when  the  plate  voltage  falls  below  the 
normal  operating  range,  the  plate  current 
falls  sharply,  while  the  screen  current  rises  to 
such  a value  that  the  total  cathode  current 
remains  substantially  constant.  Hence,  the 
screen  grid  in  a tetrode  or  pentode  will  almost 
invariably  be  damaged  by  excessive  dissipa- 
tion if  the  plate  voltage  is  removed  while  the 
screen  voltage  is  still  being  applied  from  a 
low-impedance  source. 


The  Effect  of  The  current  equations  show  how 
Grid  Current  the  total  cathode  current  in 
triodes,  tetrodes,  and  pentodes 
is  a function  of  the  potentials  applied  to  the 
various  electrodes.  If  only  one  electrode  is 
positive  with  respect  to  the  cathode  (such  as 
would  be  the  case  in  a triode  acting  as  a 
class  A amplifier)  all  the  cathode  current  goes 
to  the  plate.  But  when  both  screen  and  plate 
are  positive  in  a tetrode  or  pentode,  the  cath- 
ode current  divides  between  the  two  elements. 
Hence  the  screen  current  is  taken  from  the 
total  cathode  current,  while  the  balance  goes 
to  the  plate.  Further,  if  the  control  grid  in  a 
tetrode  or  pentode  is  operated  at  a positive 
potential  the  total  cathode  current  is  divided 
between  all  three  elements  which  have  a posi- 
tive potential.  In  a tube  which  is  receiving  a 
large  excitation  voltage,  it  may  be  said  that 
the  control  grid  robs  electrons  from  the  output 
electrode  during  the  period  that  the  grid  is 
positive,  making  it  always  necessary  to  limit 
the  peak-positive  excursion  of  the  control 
grid. 


Coefficients  of  In  general  it  may  be  stated 
Tetrodes  and  that  the  amplification  factor 
Pentodes  of  tetrode  and  pentode  tubes 

is  a coefficient  which  is  not 
of  much  use  to  the  designer.  In  fact  the  ampli- 
fication factor  is  seldom  given  on  the  design 
data  sheets  of  such  tubes.  Its  value  is  usually 
very  high,  due  to  the  relatively  high  plate 
resistance  of  such  tubes,  but  bears  little 
relationship  to  the  stage  gain  which  actually 
will  be  obtained  with  such  tubes. 

On  the  other  hand,  the  grid-plate  transcon- 
ductance is  the  most  important  coefficient  of 
pentode  and  tetrode  tubes.  Gain  per  stage  can 
be  computed  directly  when  the  is  known. 
The  grid-plate  transconductance  of  a tetrode 
or  pentode  tube  can  be  calculated  through  use 
of  the  expression: 


G„. 


AE, 


with  Egg  and  Ep  constant. 

The  plate  resistance  of  such  tubes  is  of 
less  importance  than  in  the  case  of  triodes, 
though  it  is  often  of  value  in  determining  the 
amount  of  damping  a tube  will  exert  upon  the 
impedance  in  its  plate  circuit.  Plate  resist- 
ance is  calculated  from: 


Rp 


AEp 


with  Eg  and  E^g  constant. 


4-5  Mixer  and  Converter  Tubes 


The  superheterodyne  receiver  always  in- 


80  Vacuum  Tube  Principles 


THE  RADIO 


CATHODE -H 


OSCILLATOR  GRID 

SCREEN  GRID 

PLATE 


K METAL  SHELL 


-SUPPRESSOR  AND  SHELL 
• — SIGNAL  GRID 


Figure  20 

GRID  STRUCTURE  OF  6SA7 
CONVERTER  TUBE 


eludes  at  least  one  stage  for  changing  the 
frequency  of  the  incoming  signal  to  the  fixed 
frequency  of  the  main  intermediate  amplifier 
in  the  receiver.  This  frequency  changing 
process  is  accomplished  by  selecting  the 
beat-note  difference  frequency  between  a 
locally  generated  oscillation  and  the  incoming 
signal  frequency.  If  the  oscillator  signal  is 
supplied  by  a separate  tube,  the  ftequency 
changing  tube  is  called  a mixer.  Alternatively, 
the  oscillation  may  be  generated  by  additional 
elements  within  the  ftequency  changer  tube. 
In  this  case  the  ftequency  changer  is  common- 
ly called  a converter  tube. 

Conversion  The  conversion  conductance  (0^) 
Conductonce  is  a coefficient  of  interest  in  the 
case  of  mixer  or  converter  tubes, 
or  of  conventional  triodes,  tetrodes,  or  pen- 
todes operating  as  ftequency  changers.  The 
conversion  conductance  is  the  ratio  of  a 
change  in  the  signal-grid  voltage  at  the  input 
ftequency  to  a change  in  the  output  current  at 
the  converted  ftequency.  Hence  in  a mixer 
is  essentially  the  same  as  ttansconductance 
in  an  amplifier,  with  the  exception  that  the 
input  signal  and  the  output  current  ate  on  dif- 
ferent frequencies.  The  value  of  G(.  in  con- 
ventional mixer  tubes  is  from  300  to  1000 
micromhos.  The  value  of  G(,  in  an  amplifier 
tube  operated  as  a mixer  is  approximately  0.3 
the  G,n  of  the  tube  operated  as  an  amplifier. 
The  voltage  gain  of  a mixer  stage  is  equal  to 
GcZl  where  Zl  is  the  impedance  of  the  plate 
load  into  which  the  mixer  tube  operates. 

The  Diode  Mixer  The  simplest  mixer  tube  is 
the  diode.  The  noise  figure, 
or  figure  of  merit,  for  a mixer  of  this  type  is 
not  as  good  as  that  obtained  with  other  more 
complex  mixers;  however,  the  diode  is  useful 
as  a mixer  in  u-h-f  and  v-h-f  equipment  where 
low  interelectrode  capacities  are  vital  to  cir- 
cuit operation.  Since  the  diode  impedance  is 


Figure  21 

SHOWING  THE  EFFECT  OF  CATHODE 
LEAD  INDUCTANCE 

The  degenerative  action  of  cathode  lead  In. 
ductance  tends  to  reduce  the  effective  grld-to- 
cathode  voltage  with  respect  to  the  voltage 
avollohle  across  the  Input  tuned  circuit.  Cath- 
ode lead  inductance  also  Introduces  undesir. 
able  coupling  between  the  Input  and  the  out- 
put  circuits. 


low,  the  local  oscillator  must  furnish  con- 
siderable power  to  the  diode  mixer.  A good 
diode  mixer  has  an  overall  gain  of  about  0.5. 

The  Triode  Mixer  A ttiode  mixer  has  better 
gain  and  a better  noise  figure 
than  the  diode  mixer.  At  low  frequencies,  the 
gain  and  noise  figure  of  a triode  mixer  closely 
approaches  those  figures  obtained  when  the 
tube  is  used  as  an  amplifier.  In  the  u-h-f  and 
v-h-f  range,  the  efficiency  of  the  ttiode  mixer 
deteriorates  rapidly.  The  optimum  local  oscil- 
lator voltage  for  a triode  mixer  is  about  0.7  as 
large  as  the  cutoff  bias  of  the  ttiode.  Very 
little  local  oscillator  power  is  required  by  a 
ttiode  mixer. 

Pentode  Mixers  and  The  most  common  multi- 
Converter  Tubes  grid  converter  tube  for 
broadcast  or  shortwave 
use  is  the  penta  grid  converter,  typified  by 
the  6SA7,  6SB7-Y  and  6BA7  tubes  (figure  20). 
Operation  of  these  converter  tubes  and  pentode 
mixers  will  be  coveted  in  the  Receiver  Funda- 
mentals Chapter. 


4-6  Electron  Tubes  at  Very 
High  Frequencies 


As  the  frequency  of  operation  of  the  usual 
type  of  electron  tube  is  increased  above  about 
20  Me.,  certain  assumptions  which  ate  valid 
for  operation  at  lower  frequencies  must  be  re- 
examined. First,  we  find  that  lead  inductances 
from  the  socket  connections  to  the  actual 
elements  within  the  envelope  no  longer  are 
negligible.  Second,  we  find  that  electron 


HANDBOOK 


The  Klystron  81 


transit  time  no  longer  may  be  ignored;  an 
appreciable  fraction  of  a cycle  of  input  signal 
may  be  required  for  an  electron  to  leave  the 
cathode  space  charge,  pass  through  the  grid 
wires,  and  travel  through  the  space  between 
grid  and  plate. 

Effects  of  The  effect  of  lead  induct- 

Leod  Inductance  ance  is  two-fold.  First,  as 
shown  in  figure  21,  the 
combination  of  grid-lead  inductance,  grid- 
cathode  capacitance,  and  cathode  lead  induct- 
ance tends  to  reduce  the  effective  grid-cathode 
signal  voltage  for  a constant  voltage  at  the 
tube  terminals  as  the  frequency  is  increased. 
Second,  cathode  lead  inductance  tends  to 
introduce  undesired  coupling  between  the 
various  elements  within  the  tube. 

Tubes  especially  designed  for  v-h-f  and 
u-h-f  use  have  had  their  lead  inductances 
minimized.  The  usual  procedures  for  reducing 
lead  inductance  are:  (1)  using  heavy  lead 
conductors  or  several  leads  in  parallel  (ex- 
amples are  the  6SH7  and  6AK5),  (2)  scaling 
down  the  tube  in  all  dimensions  to  reduce 
both  lead  inductances  and  interelectrode 
capacitances  (examples  are  the  6AK5,  6F4, 
and  other  acorn  and  miniature  tubes),  and  (3) 
the  use  of  very  low  inductance  extensions  of 
the  elements  themselves  as  external  connec- 
tions (examples  are  lighthouse  tubes  such  as 
the  2C40,  oilcan  tubes  such  as  the  2C29,  and 
many  types  of  v-h-f  transmitting  tubes). 

Effect  of  When  an  electron  tube  is  op- 

Tronsit  Time  erated  at  a frequency  high 
enough  that  electron  transit 
time  between  cathode  and  plate  is  an  ap- 
preciable fraction  of  a cycle  at  the  input  fre- 
quency, several  undesirable  effects  take  place. 
First,  the  grid  takes  power  from  the  input 
signal  even  though  the  grid  is  negative  at  all 
times.  This  comes  about  since  the  grid  will 
have  changed  its  potential  during  the  time 
required  for  an  electron  to  pass  from  cathode 
to  plate.  Due  to  interaction,  and  a resulting 
phase  difference  between  the  field  associated 
with  the  grid  and  that  associated  with  a mov- 
ing electron,  the  grid  presents  a resistance  to 
an  input  signal  in  addition  to  its  normal 
"cold”  capacitance.  Further,  as  a result  of 
this  action,  plate  current  no  longer  is  in  phase 
with  grid  voltage. 

An  amplifier  stage  operating  at  a frequency 
high  enough  that  transit  time  is  appreciable: 

(a)  Is  difficult  to  excite  as  a result  of  grid 
loss  from  the  equivalent  input  grid  resistance, 

(b)  Is  capable  of  less  output  since  trans- 
conductance is  reduced  and  plate  current  is 
not  in  phase  with  grid  voltage. 

The  effects  of  transit  time  increase  with  the 
square  of  the  operating  frequency,  and  they 


increase  rapidly  as  frequency  is  increased 
above  the  value  where  they  become  just  ap- 
preciable. These  effects  may  be  reduced  by 
scaling  down  tube  dimensions;  a procedure 
which  also  reduces  lead  inductance.  Further, 
transit-time  effects  may  be  reduced  by  the 
obvious  procedure  of  increasing  electrode  po- 
tentials so  that  electron  velocity  will  be  in- 
creased. However,  due  to  the  law  of  electron- 
motion  in  an  electric  field,  transit  time  is 
increased  only  as  the  square  root  of  the  ratio 
of  operating  potential  increase;  therefore  this 
expedient  is  of  limited  value  due  to  other 
limitations  upon  operating  voltages  of  small 
electron  tubes. 


4-7  Special  Microwave 

Electron  Tubes 

Due  primarily  to  the  limitation  imposed  by 
transit  time,  conventional  negative-grid  elec- 
tron tubes  are  capable  of  affording  worthwhile 
amplification  and  power  output  only  up  to  a 
definite  upper  frequency.  This  upper  frequency 
limit  varies  from  perhaps  100  Me.  for  con- 
ventional tube  types  to  about  4000  Me.  for 
specialized  types  such  as  the  lighthouse  tube. 
Above  the  limiting  frequency,  the  conventional 
negative-grid  tube  no  longer  is  practicable  and 
recourse  must  be  taken  to  totally  different 
types  of  electron  tubes  in  which  electron 
transit  time  is  not  a limitation  to  operation. 
Three  of  the  most  important  of  such  microwave 
tube  types  are  the  klystron,  the  magnetron,  and 
the  travelling  wave  tube. 

The  Power  Klystron  The  klystron  is  a type 
of  electron  tube  in  which 
electron  transit  time  is  used  to  advantage. 
Such  tubes  comprise,  as  shown  in  figure  22, 
a cathode,  a focussing  electrode,  a resonator 
connected  to  a pair  of  grids  which  afford 
velocity  modulation  of  the  electron  beam 
(called  the  "buncher”),  a drift  space,  and 
another  resonator  connected  to  a pair  of  grids 
(called  the  "catcher”).  A collector  for  the 
expended  electrons  may  be  included  at  the 
end  of  the  tube,  or  the  catcher  may  also  per- 
form the  function  of  electron  collection. 

The  tube  operates  in  the  following  manner: 
The  cathode  emits  a stream  of  electrons  which 
is  focussed  into  a beam  by  the  focussing 
electrode.  The  stream  passes  through  the 
buncher  where  it  is  acted  upon  by  any  field 
existing  between  the  two  grids  of  the  buncher 
cavity.  When  the  potential  between  the  two 
grids  is  zero,  the  stream  passes  through  with- 
out change  in  velocity.  But  when  the  potential 
between  the  two  grids  of  the  buncher  is  in- 
creasingly positive  in  the  direction  of  electron 


82  Vacuum  Tube  Principles 


THE  RADIO 


Figure  22 

TWO-CAVITY  KLYSTRON  OSCILLATOR 

A convenriono/  two-eavify  klyitron  Is  shown 
with  a feedback  loop  connected  between  the 
two  cavities  so  that  the  tube  may  be  used  as 
an  oscillator. 


motion,  the  velocity  of  the  electrons  in  the 
beam  is  increased.  Conversely,  when  the  field 
becomes  increasingly  negative  in  the  direction 
of  the  beam  (corresponding  to  the  other  half 
cycle  of  the  exciting  voltage  from  that  which 
produced  electron  acceleration)  the  velocity 
of  the  electrons  in  the  beam  is  decreased. 

When  the  velocity-modulated  electron  beam 
reaches  the  drift  space,  where  there  is  no  field, 
those  electrons  which  have  been  sped  up  on 
one  half-cycle  overtake  those  immediately 
ahead  which  were  slowed  down  on  the  other 
half-cycle.  In  this  way,  the  beam  electrons  be- 
come bunched  together.  As  the  bunched  groups 
pass  through  the  two  grids  of  the  catcher 
cavity,  they  impart  pulses  of  energy  to  these 
grids.  The  catcher  grid-space  is  charged  to 
different  voltage  levels  by  the  passing  electron 
bunches,  and  a corresponding  oscillating  field 
is  set  up  in  the  catcher  cavity.  The  catcher  is 
designed  to  resonate  at  the  frequency  of  the 
velocity-modulated  beam,  or  at  a harmonic  of 
this  frequency. 

In  the  klystron  amplifier,  energy  delivered 
by  the  buncher  to  the  catcher  grids  is  greater 
than  that  applied  to  the  buncher  cavity  by  the 
input  signal.  In  the  klystron  oscillator  a feed- 
back loop  connects  the  two  cavities.  Coupling 
to  either  buncher  or  catcher  is  provided  by 
small  loops  which  enter  the  cavities  by  way  of 
concentric  lines. 

The  klystron  is  an  electron-coupled  device. 
When  used  as  an  oscillator,  its  output  voltage 
is  rich  in  harmonics.  Klystron  oscillators  of 
various  types  afford  power  outputs  ranging 
from  less  than  1 watt  to  many  thousand  watts. 
Operating  efficiency  varies  between  5 and  30 
per  cent.  Frequency  may  be  shifted  to  some 
extent  by  varying  the  beam  voltage.  Tuning  is 


Figure  23 

REFLEX  KLYSTRON  OSCILLATOR 

A conventional  reflex  klystron  oscillator  of 
the  type  commonly  used  as  a local  oscillator 
in  superheterodyne  receivers  operating  above 
about  2000  Me.  is  shown  above.  Frequency 
modulation  of  the  output  frequency  of  the  oscil- 
lator, or  a-f-c  operation  in  a receiver,  may  be 
obtained  by  varying  the  negative  voltage  on  the 
repeller  electrode. 


carried  on  mechanically  in  some  klystrons  by 
altering  (by  means  of  knob  settings)  the  shape 
of  the  resonant  cavity. 

The  Reflex  Klystron  The  two-cavity  klystron 
as  described  in  the  pre- 
ceding paragraphs  is  primarily  used  as  atrans- 
mitting  device  since  quite  reasonable  amounts 
of  power  are  made  available  in  its  output  cir- 
cuit. However,  for  applications  where  a much 
smaller  amount  of  power  is  required  — power 
levels  in  the  milliwatt  range  — for  low-power 
transmitters,  receiver  local  oscillators,  etc., 
another  type  of  klystron  having  only  a single 
cavity  is  more  frequently  used. 

The  theory  of  operation  of  the  single-cavity 
klystron  is  essentially  the  same  as  the  multi- 
cavity type  with  the  exception  that  the  veloc- 
ity-modulated electron  beam,  after  having  left 
the  "buncher”  cavity  is  reflected  back  into 
the  area  of  the  buncher  again  by  a repeller 
electrode  as  illustrated  in  figure  23.  The 
potentials  on  the  various  electrodes  are  ad- 
justed to  the  value  such  that  proper  bunching 
of  the  electron  beam  will  take  place  just  as  a 
particular  portion  of  the  velocity-modulated 
beam  reenters  the  area  of  the  resonant  cavity. 
Since  this  type  of  klystron  has  only  one  circuit 
it  can  be  used  only  as  an  oscillator  and  not  as 
an  amplifier.  Effective  modulation  of  the  fre- 
quency of  a single-cavity  klystron  for  FM 
work  can  be  obtained  by  modulating  the  te- 
peller  electrode  voltage. 


HANDBOOK 


The  Magnetron  83 


GRID  ANODE  ANODE 


Figure  24 

CUTAWAY  VIEW  OF 
WESTERN  ELECTRIC  416-B/6280 
VHF  PLANAR  TRIODE  TUBE 
The  4 76*6,  designed  by  the  Bell 
Telephone  Loboratories  is  intended 
for  amplifier  or  frequency  multiplier 
service  in  the  4000  me  reg/on.  Em- 
ploying  grid  wires  having  a diameter 
equal  to  fifteen  wavelengths  of  fight, 
the  4 76*6  has  a transconductance  of 
50,000.  Spacing  between  grid  and 
cathode  is  .0005",  to  reduce  transit 
time  effects.  Entire  tube  is  gold  plated. 


The  Magnetron  The  magnetron  is  an  s-h-f 
oscillator  tube  normally  em- 
ployed where  very  high  values  of  peak  power 
or  moderate  amounts  of  average  power  are 
required  in  the  range  from  perhaps  700  Me. 
to  30,000  Me.  Special  magnetrons  were  de- 
veloped for  wartime  use  in  radar  equipments 
which  had  peak  power  capabilities  of  several 
million  watts  (megawatts)  output  at  frequen- 
cies in  the  vicinity  of  3000  Me.  The  normal 
duty  cycle  of  operation  of  these  radar  equip- 
ments was  approximately  1/10  of  one  per 
cent  (the  tube  operated  about  1/1000  of  the 
time  and  rested  for  the  balance  of  the  operat- 
ing period)  so  that  the  average  power  output 

of  these  magnetrons  was  in  the  vicinity  of 
1000  watts. 


® 


Figure  25 

SIMPLE  MAGNETRON  OSCILLATOR 

An  external  iank  circuit  It  uted  with  this  type 
of  magnetron  oscillator  for  operation  In  the 
lower  U‘h*f  range. 


In  its  simplest  form  the  magnetron  tube  is  a 
filament-type  diode  with  two  half-cylindrical 
plates  or  anodes  situated  coaxially  with  re- 
spect to  the  filament.  The  construction  is 
illustrated  in  figure  25A.  The  anodes  of  the 
magnetron  are  connected  to  a resonant  circuit 
as  illustrated  on  figure  25B.  The  tube  is  sur- 
rounded by  an  electromagnet  coil  «fiich,  in 
turn,  is  connected  to  a low-voltage  d-c  ener- 
gizing source  through  a rheostat  R for  control- 
ling the  strength  of  the  magnetic  field.  The 
field  coil  is  oriented  so  that  the  lines  of 
magnetic  force  it  sets  up  are  parallel  to  the 
axis  of  the  electrodes. 

Under  the  influence  of  the  strong  magnetic 
field,  electrons  leaving  the  filament  are  de- 
flected from  their  normal  paths  and  move  in 
circular  orbits  within  the  anode  cylinder.  This 
effect  results  in  a negative  resistance  which 
sustains  oscillations.  The  oscillation  fre- 
quency is  very  nearly  the  value  determined  by 
L and  C.  In  other  magnetron  circuits,  the  fre- 
quency may  be  governed  by  the  electron  rota- 
tion, no  external  tuned  circuits  being  em- 
ployed. Wavelengths  of  less  than  1 centi- 
meter have  been  produced  with  such  circuits. 

More  complex  magnetron  tubes  employ  no 
external  tuned  circuit,  but  utilize  instead  one 
or  more  resonant  cavities  which  are  integral 
with  the  anode  structure.  Figure  26  shows  a 
magnetron  of  this  type  having  a multi-cellular 


84  Vacuum  Tube  Principles 


THE  RADIO 


Figure  26 

MODERN  MULTI-CAVITY  MAGNETRON 

Illustrated  Is  an  exiernal-anede  strapped  mag- 
netron of  the  type  common//  used  In  radar  equip- 
ment for  the  10-cm.  range.  A permanent  magnet 
of  the  genera/  type  used  with  such  a magnetron 
Is  shown  In  the  right-hand  port/on  of  the  (/rowing, 
with  the  magnetron  in  place  between  the  pole 
pieces  of  the  magnet. 


anode  of  eight  cavities.  It  will  be  noted,  also, 
that  alternate  cavities  (which  would  operate  at 
the  same  polarity  when  the  tube  is  oscillating) 
are  strapped  together.  Strapping  was  found  to 
improve  the  efficiency  and  stability  of  high- 
power  radar  magnetrons.  In  most  radar  appli- 
cations of  magnetron  oscillators  a powerful 
permanent  magnet  of  controlled  characteristics 
is  employed  to  supply  the  magnetic  field 
rather  than  the  use  of  an  electromagnet. 

The  Travelling  The  Travelling  'Have  Tube 
Wave  Tube  (figure  27)  consists  of  a helix 
located  within  an  evacuated 
envelope.  Input  and  output  terminations  are 
affixed  to  each  end  of  the  helix.  An  electron 
beam  passes  through  the  helix  and  interacts 
with  a wave  travelling  along  the  helix  to  pro- 
duce broad  band  amplification  at  microwave 
frequencies. 

When  the  input  signal  is  applied  to  the  gun 
end  of  the  helix,  it  travels  along  the  helix  wire 
at  approximately  the  speed  of  light.  However, 
the  signal  velocity  measured  along  the  axis 
of  the  helix  is  considerably  lower.  The  elec- 
trons emitted  by  the  cathode  gun  pass  axially 
through  the  helix  to  the  collector,  located  at 
the  output  end  of  the  helix.  The  average  veloc- 
ity of  the  electrons  depends  upon  the  potential 
of  the  collector  with  respect  to  the  cathode. 
When  the  average  velocity  of  the  electrons  is 
greater  than  the  velocity  of  the  helix  wave, 
the  electrons  become  crowded  together  in  the 
various  regions  of  retarded  field,  where  they 
impart  energy  to  the  helix  wave.  A power  gain 
of  100  or  more  may  be  produced  by  this  tube. 

4-8  The  Cathode-Ray  Tube 

The  Cathode- Ray  Tube  The  cathode-ray  tube 
is  a special  type  of 


WAVEGUIDE  WAVEGUIDE 

INPUT  OUTPUT 


Operation  of  this  tube  is  the  result  of  inter- 
action between  the  electron  beam  and  wave 
travelling  along  the  helix. 


electron  tube  which  permits  the  visual  observa- 
tion of  electrical  signals.  It  may  be  incorpo- 
rated into  an  oscilloscope  for  use  as  a test 
instrument  or  it  may  be  the  display  device  for 
radar  equipment  or  a television  receiver. 

Operation  of  A cathode-ray  tube  always  in- 
the  CRT  eludes  an  electron  gun  for  pro- 
ducing a stream  of  electrons,  a 
grid  for  controlling  the  intensity  of  the  elec- 
tron beam,  and  a luminescent  screen  for  con- 
verting the  impinging  electron  beam  into  visi- 
ble light.  Such  a tube  always  operates  in  con- 
junction with  either  a built-in  or  an  external 
means  for  focussing  the  electron  stream  into  a 
narrow  beam,  and  a means  for  deflecting  the 
electron  beam  in  accordance  with  an  electrical 
signal. 

The  main  electrical  difference  between 
types  of  cathode-ray  tubes  lies  in  the  means 
employed  for  focussing  and  deflecting  the 
electron  beam.  The  beam  may  be  focussed 
and/or  deflected  either  electrostatically  or 
magnetically,  since  a stream  of  electrons  can 
be  acted  upon  either  by  an  electrostatic  or  a 
magnetic  field.  In  an  electrostatic  field  the 
electron  beam  tends  to  be  deflected  toward  the 
positive  termination  of  the  field  (figure  28). 
In  a magnetic  field  the  stream  tends  to  be 
deflected  at  right  angles  to  the  field.  Further, 
an  electron  beam  tends  to  be  deflected  so  that 
it  is  normal  (perpendicular)  to  the  equipotential 
lines  of  an  electrostatic  field — andittendsto 
be  deflected  so  that  it  is  parallel  to  the  lines 
of  force  in  a magnetic  field. 

Large  cathode-ray  tubes  used  as  kinescopes 
in  television  receivers  usually  are  both  focused 
and  deflected  magnetically.  On  the  other  hand, 
the  medium-size  CR  tubes  used  in  oscillo- 
scopes and  small  television  receivers  usually 
are  both  focused  and  deflected  electrostat- 
ically. But  CR  tubes  for  special  applications 
may  be  focused  magnetically  and  deflected 
electrostatically  or  vice  versa. 

There  are  advantages  and  disadvantages  to 


HANDBOOK 


The  Cathode  Ray  Tube  85 


Figure  28 

TYPICAL  ELECTROSTATIC 
CATHODE. RAY  TUBE 


anode,  A,  which  is  operated  at  a high  positive 
potential.  In  some  tubes  this  electrode  is  oper- 
ated at  a higher  potential  than  the  first  accel- 
erating electrode,  H,  while  in  other  tubes  both 
accelerating  electrodes  are  operated  at  the 
same  potential. 

The  electrodes  which  have  been  described 
up  to  this  point  constitute  the  electron  gun, 
which  produces  the  free  electrons  and  focusses 
them  into  a slender,  concentrated,  rapidly- 
traveling  stream  for  projecting  onto  the  view- 
ing screen. 


both  types  of  focussing  and  deflection.  How- 
ever, it  may  be  stated  that  electrostatic  deflec- 
tion is  much  better  than  magnetic  deflection 
when  high-frequency  waves  are  to  be  displayed 
on  the  screen;  hence  the  almost  universal  use 
of  this  type  of  deflection  for  oscillographic 
work.  But  when  a tube  is  operated  at  a high 
value  of  accelerating  potential  so  as  to  obtain 
a bright  display  on  the  face  of  the  tube  as  for 
television  or  radar  work,  the  use  of  magnetic 
deflection  becomes  desirable  since  it  is  rela- 
tively easier  to  deflect  a high-velocity  electron 
beam  magnetically  than  electrostatically. 
However,  an  ion  trap  is  required  with  mag- 
netic deflection  since  the  heavy  negative  ions 
emitted  by  the  cathode  ate  not  materially  de- 
flected by  the  magnetic  field  and  hence  would 
burn  an  "ion  spot’’  in  the  center  of  the  lumi- 
nescent screen.  With  electrostatic  deflection 
the  heavy  ions  are  deflected  equally  as  well 
as  the  electrons  in  the  beam  so  that  an  ion 
spot  is  not  formed. 

Construction  of  The  construction  of  atypical 
Electrostatic  CRT  electrostatic-focus,  electro- 
static-deflection cathode- ray 
tube  is  illustrated  in  the  pictorial  diagram  of 
figure  28.  The  indirectly  heated  cathode  K re- 
leases free  electrons  when  heated  by  the 
enclosed  filament.  The  cathode  is  surrounded 
by  a cylinder  G,  which  has  a small  hole  in  its 
front  for  the  passage  of  the  electron  stream. 
Although  this  element  is  not  a wire  mesh  as 
is  the  usual  grid,  it  is  known  by  the  same 
name  because  its  action  is  similar;  it  controls 
the  electron  stream  when  its  negative  potential 
is  varied. 

Next  in  order,  is  found  the  first  accelerating 
anode,  H,  which  resembles  another  disk  or 
cylinder  with  a small  hole  in  its  center.  This 
electrode  is  run  at  a high  or  moderately  high 
positive  voltage,  to  accelerate  the  electrons 
towards  the  far  end  of  the  tube. 

The  focussing  electrode,  F,  is  a sleeve 
which  usually  contains  two  small  disks,  each 
with  a small  hole. 

After  leaving  the  focussing  electrode,  the 
electrons  pass  through  another  accelerating 


Electrostatic  To  make  the  tube  useful,  means 
Deflection  must  be  provided  for  deflecting 
the  electron  beam  along  two  axes 
at  right  angles  to  each  other.  The  more  com- 
mon tubes  employ  electrostatic  deflection 
plates,  one  pair  to  exert  a force  on  the  beam 
in  the  vertical  plane  and  one  pair  to  exert  a 
force  in  the  horizontal  plane.  These  plates 
are  designated  as  B and  C in  figure  28. 

Standard  oscilloscope  practice  with  small 
cathode-ray  tubes  calls  for  connecting  one  of 
the  B plates  and  one  of  the  C plates  together 
and  to  the  high  voltage  accelerating  anode. 
With  the  newer  three-inch  tubes  and  with  five- 
inch  tubes  and  larger,  all  four  deflecting  plates 
ate  commonly  used  for  deflection.  The  positive 
high  voltage  is  grounded,  instead  of  the  nega- 
tive as  is  common  practice  in  amplifiers,  etc., 
in  order  to  permit  operation  of  the  deflecting 
plates  at  a d-c  potential  at  or  near  ground. 

An  Aquadag  coating  is  applied  to  the  inside 
of  the  envelope  to  attract  any  secondary  elec- 
trons emitted  by  the  flourescent  screen. 

In  the  average  electrostatic-deflection  CR 
tube  the  spot  will  be  fairly  well  centered  if  all 
four  deflection  plates  ate  returned  to  the  po- 
tential of  the  second  anode  (ground).  How- 
ever, for  accurate  centering  and  to  permit  mov- 
ing the  entire  trace  either  horizontally  or 
vertically  to  permit  display  of  a particular 
waveform,  horizontal  and  vertical  centering 
controls  usually  are  provided  on  the  front  of 
the  oscilloscope. 

After  the  spot  is  once  centered,  it  is  neces- 
sary only  to  apply  a positive  or  negative  volt- 
age (with  respect  to  ground)  to  one  of  the 
ungrounded  or  "free’’  deflector  plates  in  order 
to  move  the  spot.  If  the  voltage  is  positive 
with  respect  to  ground,  the  beam  will  be 
attracted  toward  that  deflector  plate,  while  if 
negative  the  beam  and  spot  will  be  repulsed. 
The  amount  of  deflection  is  directly  propor- 
tional to  the  voltage  (with  respect  to  ground) 
that  is  applied  to  the  free  electrode. 

With  the  larget-screen  higher-voltage  tubes 
it  becomes  necessary  to  place  deflecting  volt- 
age on  both  horizontal  and  both  vertical  plates. 
This  is  done  for  two  reasons:  First,  the  amount 
of  deflection  voltage  required  by  the  high- 


86  Vacuum  Tube  Principles 


THE  RADIO 


Figure  29 

TYPICAL  ELECTROMAGNETIC 
CATHODE-RAY  TUBE 


voltage  tubes  is  so  great  that  a transmitting 
tube  operating  from  a high  volrage  supply 
would  be  required  to  attain  this  voltage  with- 
out distortion.  By  using  push-pull  deflection 
with  two  tubes  feeding  the  deflection  plates, 
the  necessary  plate  supply  voltage  for  the  de- 
flection amplifier  is  halved.  Second,  a certain 
amount  of  de-focussing  of  the  electron  stream 
is  always  present  on  the  extreme  excursions  in 
deflection  voltage  when  this  voltage  is  applied 
only  to  one  deflecting  plate.  When  the  de- 
flecting voltage  is  fed  in  push-pull  to  both 
deflecting  plates  in  each  plane,  there  is  no  de- 
focussing because  the  average  voltage  acting 
on  the  electron  stream  is  zero,  even  though  the 
net  voltage  (which  causes  the  deflection) 
acting  on  the  stream  is  twice  that  on  either 
plate. 

The  fact  that  the  beam  is  deflected  by  a 
magnetic  field  is  important  even  in  an  oscillo- 
scope which  employs  a tube  using  electro- 
static deflection,  because  it  means  that  pre- 
cautions must  be  taken  to  protect  the  tube  from 
the  transformer  fields  and  sometimes  even  the 
earth’s  magnetic  field.  This  normally  is  done 
by  incorporating  a magnetic  shield  around  the 
tube  and  by  placing  any  transformers  as  far 
from  the  tube  as  possible,  oriented  to  the  posi- 
tion which  produces  minimum  effect  upon  the 
electron  stream. 

Construction  of  Electro-  The  electromagnetic 
magnetic  CRT  cathode-ray  tube  allows 

greater  definition  than 
does  the  electrostatic  tube.  Also,  electro- 
magnetic definition  has  a number  of  advan- 
tages when  a rotating  radial  sweep  is  required 
to  give  polar  indicarions. 

The  producrion  of  the  electron  beam  in  an 
electromagnetic  tube  is  essentially  the  same 
as  in  the  electrostatic  tube.  The  grid  structure 
is  similar,  and  controls  the  electron  beam  in 
an  identical  manner.  The  elements  of  a typical 
electromagnetic  tube  ate  shown  in  figure  29. 
The  focus  coil  is  wound  on  an  iron  cote  which 
may  be  moved  along  the  neck  of  the  tube  to 
focus  the  electron  beam.  For  final  adjustment. 


Figure  30 

Two  pairs  of  coils  arranged  for  electro- 
magnetic deflection  in  two  directions. 


the  current  flowing  in  the  coil  may  be  varied. 
A second  pair  of  coils,  the  deflection  coils 
are  mounted  at  right  angles  to  each  other 
around  the  neck  of  the  tube.  In  some  cases, 
these  coils  can  rotate  around  the  axis  of  the 
tube. 

Two  anodes  ate  used  for  accelerating  the 
electrons  from  the  cathode  to  the  screen.  The 
second  anode  is  a graphite  coating  (Aquadag) 
on  the  inside  of  the  glass  envelope.  The  func- 
tion of  this  coating  is  to  attract  any  secondary 
electrons  emitted  by  the  floutescent  screen, 
and  also  to  shield  the  electron  beam. 

In  some  types  of  electromagnetic  tubes,  a 
first,  or  accelerating  anode  is  also  used  in 
addition  to  the  Aquadag. 

Electromagnetic  A magnetic  field  will  deflect 
Deflection  an  electron  beam  in  a direc- 

tion which  is  at  right  angles 
to  both  the  direction  of  the  field  and  the  direc- 
tion of  motion  of  the  beam. 

In  the  general  case,  two  pairs  of  deflection 
coils  are  used  (figure  30).  One  pair  is  for 
horizontal  deflection,  and  the  other  pair  is  for 
vertical  deflection.  The  two  coils  in  a pair 
are  connected  in  series  and  are  wound  in  such 
directions  that  the  magnetic  field  flows  from 
one  coil,  through  the  electron  beam  to  the 
other  coil.  The  force  exerted  on  the  beam  by 
the  field  moves  it  to  any  point  on  the  screen 
by  application  of  the  proper  currents  to  these 
coils. 

The  Trace  The  human  eye  retains  an  image 
for  about  one- sixteenth  second 
after  viewing.  In  a CRT,  rhe  spot  can  be 
moved  so  quickly  that  a series  of  adjacent 
spots  can  be  made  to  appear  as  a line,  if  the 
beam  is  swept  over  the  path  fast  enough.  As 


HANDBOOK 


Gas  Tubes  87 


long  as  the  electron  beam  strikes  in  a given 
place  at  least  sixteen  times  a second,  the 
spot  will  appear  to  the  human  eye  as  a source 
of  continuous  light  with  very  little  flicker. 

Screen  Materials  — At  least  five  types  of  lumi- 
“Phosphors*'  nescent  screen  materials 

are  commonly  available  on 
the  various  types  of  CR  tubes  commercially 
available.  These  screen  materials  are  called 
phosphors;  each  of  the  five  phosphors  is  best 
suited  to  a particular  type  of  application.  The 
P‘l  phosphor,  which  has  a green  flourescence 
with  medium  persistence,  is  almost  invariably 
used  for  oscilloscope  tubes  for  visual  observa- 
tion. The  P-4  phosphor,  with  white  fluores- 
cence and  medium  persistence,  is  used  on 
television  viewing  tubes  ("Kinescopes^*).  The 
P-5  and  P-11  phosphors,  with  blue  fluores- 
cence and  very  short  persistence,  are  used 
primarily  in  oscilloscopes  where  photographic 
recording  of  the  trace  is  to  be  obtained.  The 
P-7  phosphor,  which  has  a blue  flash  and  a 
long-persistence  greenish-yellow  persistence, 
is  used  primarily  for  radar  displays  where 
retention  of  the  image  for  several  seconds 
after  the  initial  signal  display  is  required. 

4-9  Gas  Tubes 


The  space  charge  of  electrons  in  the  vicinity 
of  the  cathode  in  a diode  causes  the  plate-to- 
cathode  voltage  drop  to  be  a function  of  the 
current  being  carried  between  the  cathode  and 
the  plate.  This  voltage  drop  can  be  rather  high 
when  large  currents  are  being  passed,  causing 
a considerable  amount  of  energy  loss  which 
shows  up  as  plate  dissipation. 

Action  of  The  negative  space  charge  can 

Positive  Ions  be  neutralized  by  the  presence 
of  the  proper  density  of  positive 
ions  in  the  space  between  the  cathode  and 
anode.  The  positive  ions  may  be  obtained  by 
the  introduction  of  the  proper  amount  of  gas  or 
a small  amount  of  mercury  into  the  envelope  of 
the  tube.  When  the  voltage  drop  across  the 
tube  reaches  the  ionization  potential  of  the 
gas  or  mercury  vapor,  the  gas  molecules  will 
become  ionized  to  form  positive  ions.  The 
positive  ions  then  tend  to  neutralize  the  space 
charge  in  the  vicinity  of  the  cathode.  The  volt- 
age drop  across  the  tube  then  remains  constant 
at  the  ionization  potential  of  the  gas  up  to  a 
current  drain  equal  to  the  maximum  emission 
capability  of  the  cathode.  The  voltage  drop 
varies  between  10  and  20  volts,  depending 
upon  the  particular  gas  employed,  up  to  the 
maximum  current  rating  of  the  tube. 


Mercury  Vapor  Mercury-vapor  tubes,  although 
Tubes  very  widely  used,  have  the 

disadvantage  that  they  must  be 
operated  within  a specific  temperature  range 
(25°  to  70°  C.)  in  order  that  the  mercury  vapor 
pressure  within  the  tube  shall  be  within  the 
proper  range.  If  the  temperature  is  too  low, 
the  drop  across  the  tube  becomes  too  high 
causing  immediate  overheating  and  possible 
damage  to  the  elements.  If  the  temperature  is 
too  high,  the  vapor  pressure  is  too  high,  and 
the  voltage  at  which  the  tube  will  "flash  back” 
is  lowered  to  the  point  where  destruction  of 
the  tube  may  take  place.  Since  the  ambient 
temperature  range  specified  above  is  within 
the  normal  room  temperature  range,  no  trouble 
will  be  encountered  under  normal  operating 
conditions.  However,  by  the  substitution  of 
xenon  gas  for  mercury  it  is  possible  to  pro- 
duce a rectifier  with  characteristics  comparable 
to  those  of  the  mercury-vapor  tube  except  that 
the  tube  is  capable  of  operating  over  the  range 
from  approximately  —70°  to  90°  C.  The  3B25 
rectifier  is  an  example  of  this  type  of  tube. 

Thyrafron  If  a grid  is  inserted  between  the  ca- 
Tubes  thode  and  plate  of  a mercury- vapor 
gaseous-conduction  rectifier,  a neg- 
ative potential  placed  upon  the  added  element 
will  increase  the  plate-to-cathode  voltage  drop 
required  before  the  tube  will  ionize  or  "fire.” 
The  potential  upon  the  control  grid  will  have 
no  effect  on  the  plate-to-cathode  drop  after  the 
tube  has  ionized.  However,  the  grid  voltage 
may  be  adjusted  to  such  a value  that  conduc- 
tion will  take  place  only  over  the  desired 
portion  of  the  cycle  of  the  a-c  voltage  being 
impressed  upon  the  plate  of  the  rectifier. 

Voltage  Regulator  In  a glow-discharge  gas  tube 
^“hes  the  voltage  drop  across  the 

electrodes  remains  constant 
over  a wide  range  of  current  passing  through 
the  tube.  This  property  exists  because  the 
degree  of  ionization  of  the  gas  in  the  tube 
varies  with  the  amount  of  current  passing 
through  the  tube.  When  a large  current  is 
passed,  the  gas  is  highly  ionized  and  the 
internal  impedance  of  the  tube  is  low.  When  a 
small  current  is  passed,  the  gas  is  lightly 
ionized  and  the  internal  impedance  of  the  tube 
is  high.  Over  the  operating  range  of  the  tube, 
the  product  (IR)  of  the  current  through  the  tube 
and  the  internal  impedance  of  the  tube  is  very 
nearly  constant.  Examples  of  this  type  of  tube 
are  VR-150,  VR-105  and  the  old  874. 

Vacuum  Tube  Vacuum  tubes  ate  grouped  into 
Classification  three  major  classifications: 

commercial,  ruggedized,  and 
premium  (or  reliable).  Any  one  of  these  three 
groups  may  also  be  further  classified  for 


88  Vacuum  Tube  Principles 


the  radio 


military  duty  (JAN  classification).  To  qualify 
for  JAN  classification,  sample  lots  of  the 
particular  tube  must  have  passed  special 
qualification  tests  at  the  factory.  It  should  not 
be  construed  that  a JAN-type  tube  is  better 
than  a commercial  tube,  since  some  commercial 
tests  and  specifications  are  more  rigid  than 
the  corresponding  JAN  specifications.  The 
JAN- stamped  tube  has  merely  been  accepted 
under  a certain  set  of  conditions  for  military 
service. 


Ruggedized  or  Radio  tubes  ate  being  used  in 
Premium  Tubes  increasing  numbers  for  indus- 
trial applications,  such  as 
computing  and  control  machinery,  and  in  avia- 
tion and  marine  equipment.  When  a tube  fails 
in  a home  radio  receiver,  it  is  merely  incon- 
venient, but  a tube  failure  in  industrial  appli- 
cations may  bring  about  stoppage  of  some  vital 
process,  resulting  in  financial  loss,  or  even 
danger  to  life. 

To  meet  the  demands  of  these  industrial 
applications,  a series  of  tubes  was  evolved 
incorporating  many  special  features  designed 
to  ensure  a long  and  pte-detetmined  operating 
life,  and  uniform  characteristics  among  similar 
tubes.  Such  tubes  are  known  as  ruggedized  or 
premium  tubes.  Early  attempts  to  select  re- 


TRIODE  PLATE  —i pFLUORESCENT  ANODE 

TRIODE  GRID — ray  CONTROL 

\3_T_V  ELECTRODE 

CATHODES 

Figure  31 

SCHEMATIC  representation 
OF  "MAGIC  EYE”  TUBE 


liable  specimens  of  tubes  trom  ordinary  stock 
tubes  proved  that  in  the  long  tun  the  selected 
tubes  were  no  better  than  tubes  picked  at 
random.  Long  life  and  tuggedness  had  to  be 
built  into  the  tubes  by  means  of  proper  choice 
and  100%  inspection  of  all  materials  used  in 
the  tube,  by  critical  processing  inspection  and 
assembling,  and  by  conservative  ratings  of  the 
tube. 

Pure  tungsten  wire  is  used  for  heaters  in 
preference  to  alloys  of  lower  tensile  strength. 
Nickel  tubing  is  employed  around  the  heater 
wires  at  the  junction  to  the  stem  wires  to 
reduce  breakage  at  this  point.  Element  struc- 
tures are  given  extra  supports  and  bracing. 
Finally,  all  tubes  ate  given  a 50  hour  test  run 
under  full  operating  conditions  to  eliminate 
early  failures.  When  operated  within  their 
ratings,  ruggedized  or  premium  tubes  should 
provide  a life  well  in  excess  of  10,000  hours. 

Ruggedized  tubes  will  withstand  severe 
impact  shocks  for  short  periods,  and  will 


Figure  32 

amplification  factorof  typical  mode 

TUBE  DROPS  RAPIDLY  AS  PLATE  VOLTAGE 
IS  DECREASED  BELOW  20  VOLTS 

operate  under  conditions  of  vibration  for  many 
hours.  The  tubes  may  be  identified  in  many 
cases  by  the  fact  that  their  nomenclature  in- 
cludes a "W”  in  the  type  number,  as  in  807W, 
5U4W,  etc.  Some  ruggedized  tubes  are  included 
in  the  "5000”  series  nomenclature.  The  5654 
is  a ruggedized  version  of  the  6AK5,  the  5692 
is  a ruggedized  version  of  the  6SN7,  etc. 

4-10  Miscellaneous  Tube  Types 

Electron  The  electron-ray  tube  or  magic  eye 
Roy  Tubes  contains  two  sets  of  elements,  one 
of  which  is  a triode  amplifier  and 
the  other  a cathode-ray  indicator.  The  plate  of 
the  triode  section  is  internally  connected  to 
the  ray-control  electrode  (figure  31))  so  that 
as  the  plate  voltage  varies  in  accordance  with 
the  applied  signal  the  voltage  on  the  ray-control 
electrode  also  varies.  The  ray-control  electrode 
is  a metal  cylinder  so  placed  relative  to  the 
cathode  that  it  deflects  some  of  the  electrons 
emitted  from  the  cathode.  The  electrons  which 
strike  the  anode  cause  it  to  fluoresce,  or  give 
off  light,  so  that  the  deflection  caused  by  the 
ray-control  electrode,  which  prevents  electrons 
from  striking  part  of  the  anode,  produces  a 
wedge-shaped  electrical  shadow  on  the  fluores- 
cent anode.  The  size  of  this  shadow  is  deter- 
mined by  the  voltage  on  the  ray-electrode.  When 
this  electrode  is  at  the  same  potential  as  the 
fluorescent  anode,  the  shadow  disappears;  if 
the  ray-electrode  is  less  positive  than  the 
anode,  a shadow  appears  the  width  of  which 
is  proportional  to  the  voltage  on  the  ray-elec- 
trtxie.  Magic  eye  tubes  may  be  used  as  tuning 
indicators,  and  as  balance  indicators  in  electri- 
cal bridge  circuits.  If  the  angle  of  shadow  is 
calibrated,  the  eye  tube  may  be  used  as  a volt- 
meter where  rough  measurements  suffice. 


HANDBOOK 


Miscellaneous  89 


Controlled  Series  heater  strings  are  employed 
Warm-up  in  ac-dc  radio  receivers  and  tele- 
Tubes  vision  sets  to  reduce  the  cost, 

size,  and  weight  of  the  equipment. 
Voltage  surges  of  great  magnitude  occur  in 
series  operated  filaments  because  of  variations 
in  the  rate  of  warm-up  of  the  various  tubes. 
As  the  tubes  warm  up,  the  heater  resistance 
changes.  This  change  is  not  the  same  between 
tubes  of  various  types,  or  even  between  tubes 
of  the  same  type  made  by  different  manu- 
facturers. Some  6-volt  tubes  show  an  initial 
surge  as  high  as  9-volts  during  warm-up,  while 
slow-heating  tubes  such  as  the  25BQ6  are 
underheated  during  the  voltage  surge  on  the 
6-volt  tubes. 

Standardization  of  heater  characteristics  in 
a new  group  of  tubes  designed  for  series  heater 
strings  has  eliminated  this  trouble.  The  new 
tubes  have  either  600  ma.  or  400  ma.  heaters, 
with  a controlled  warm-up  time  of  approximately 
11  seconds.  The  5U8,  6CG7,  and  12BH7-A  are 
examples  of  controlled  warm-up  tubes. 

Low  Introduction  of  the  12-volt  ignition 

Plate  system  in  American  automobiles 

Potential  has  brought  about  the  design  of  a 
Tubes  series  of  tubes  capable  of  operation 

with  a plate  potential  of  12-14 
volts.  Standard  tubes  perform  poorly  at  low 
plate  potentials,  as  the  amplification  factor 
of  the  tube  drops  rapidly  as  the  plate  voltage 
is  decreased  (figure  32).  Contact  potential 
effects,  and  change  of  characteristics  with 
variations  of  filament  voltage  combine  to  make 
operation  at  low  plate  potentials  even  more 
erratic. 

By  employing  special  processing  techniques 


and  by  altering  the  electrode  geometry  a series 
of  low  voltage  tubes  has  been  developed  by 
Tung-Sol  that  effectively  perform  with  all  elec- 
trodes energized  by  a 12-volt  system.  With  a 
suitable  power  output  transistor,  this  makes 
possible  an  automobile  radio  without  a vibrator 
power  supply.  A special  space-charge  tube 
U2K5)  has  been  developed  that  delivers  40 
milliwatts  of  audio  power  with  a 12  volt  plate 
supply  (figure  33). 

Foreign  The  increased  number  of  imported 

Tubes  radios  and  high-fidelity  equipment 

have  brought  many  foreign  vacuum 
tubes  into  the  United  States.  Many  of  these 
tubes  are  comparable  to,  or  interchangeable 
with  standard  American  tubes.  A complete 
listing  of  the  electrical  characteristics  and 
base  connection  diagrams  of  all  general-pur- 
pose tubes  made  in  all  tube— producing  coun- 
tries outside  the  ’'Iron  Curtain"  is  contained 
in  the  Radio  T.ube  Vade  Mecum  (World’s  Radio 
Tubes)  available  at  most  larger  radio  parts 
jobbers  for  $5.00,  or  by  mail  from  the  publish- 
ers of  this  Handbook  at  $5.50,  postpaid.  The 
Equivalent  Tubes  Vade  Mecum  (World’s  Equiv- 
alent Tubes)  available  at  the  same  prices 
gives  all  replacement  tubes  for  a given  type, 
both  exact  and  near- equivalents  (with  points 
of  difference  detailed).  (Data  on  TV  and  spe- 
cial-purpose tubes  if  needed  is  contained  in 
a companion  volume  Television  Tubes  Vade 
Siecum). 


CHAPTER  FIVE 


Transistors  and 
Semi-Conductors 


One  of  the  earliest  detection  devices  used 
in  radio  was  the  galena  crystal,  a crude  ex- 
ample of  a semiconductor . More  modern  ex- 
amples of  semiconductors  are  the  copper- 
oxide  rectifier,  the  selenium  rectifier  and  the 
germanium  diode.  All  of  these  devices  offer 
the  interesting  property  of  greater  resistance 
to  the  flow  of  electrical  current  in  one  direc- 
tion than  in  the  opposite  direction.  Typical 
conduction  curves  for  these  semiconductors 
are  shown  in  Figure  1.  The  copper  oxide  recti- 
fier action  results  from  the  function  of  a thin 
film  of  cuprous  oxide  formed  upon  a pure  cop- 
per disc.  This  film  offers  low  resistance  for 
positive  voltages,  and  high  resistance  for 
negative  voltages.  The  same  action  is  ob- 
served in  selenium  rectifiers,  where  a film  of 
selenium  is  deposited  on  an  iron  surface. 


VOLTS 


Figure  1A 

TYPICAL  CHARACTERISTIC  CURVE 
OF  SEMI-CONDUCTOR  DIODE 


5-1  Atomic  Structure  of 
Germanium  and  Silicon 

It  has  been  previously  stated  that  the  elec- 
trons in  an  element  having  a large  atomic 
number  are  grouped  into  rings,  each  ring  hav- 
ing a definite  number  of  electrons.  Atoms  in 
which  these  rings  are  completely  filled  are 
called  inert  gases,  of  which  helium  and  argon 
are  examples.  All  other  elements  have  one  or 
more  incomplete  tings  of  electrons.  If  the  in- 
complete ring  is  loosely  bound,  the  electrons 
may  be  easily  removed,  the  element  is  called 
metallic,  and  is  a conductor  of  electric  current. 
If  the  incomplete  ring  is  tightly  bound,  with 
only  a few  missing  electrons,  the  element  is 
called  non-metallic  and  is  an  insulator  of  elec- 
tric current.  Germanium  and  silicon  fall  be- 
tween these  two  sharply  defined  groups,  and 
exhibit  both  metallic  and  non-metallic  char- 
acteristics. Pure  germanium  or  silicon  may  be 
considered  to  be  a good  insulator.  The  addition 
of  certain  impurities  in  carefully  controlled 
amounts  to  the  pure  germanium  will  alter  the 
conductivity  of  the  material.  In  addition,  the 
choice  of  the  impurity  can  change  the  direction 
of  conductivity  through  the  crystal,  some  im- 
purities increasing  conductivity  to  positive  volt- 
ages, and  others  increasing  conductivity  to  neg- 
ative voltages. 

5-2  Mechanism  of 
Conduction 

As  indicated  by  their  name,  semiconductors 
are  substances  which  have  a conductivity 
intermediate  between  the  high  values  observed 
for  metals  and  the  low  values  observed  for  in- 
sulating materials.  The  mechanism  of  conduc- 
tion in  semiconductors  is  different  from  that 


90 


Transistors  91 


ANODES 


SCHEMATIC  REPRESENTATION 


cQa 


V CATHODES 


TUBE.  GERMANIUM,  SIUCON 
AND  SEL|{N1UM  DIODES 


Figure  1-B 

COMMON  DIODE  COLOR  CODES 
AND  MARKINGS  ARE  SHOWN 
IN  ABOVE  CHART 


observed  in  metallic  conductors.  There  exist 
in  semiconductors  both  negatively  charged 
electrons  and  positively  charged  particles, 
called  holes,  which  behave  as  though  they 
had  a positive  electrical  charge  equal  in  mag- 
nitude to  the  negative  electrical  charge  on 
the  electron.  These  holes  and  electrons  drift 
in  an  electrical  field  with  a velocity  which  is 
proportional  to  the  field  itself; 

Vdh  — ghE 

where  Vdi.  = drift  velocity  of  hole 

E = magnitude  of  electric  field 
gh  = mobility  of  hole 

In  an  electric  field  the  holes  will  drift  in  a 
direction  opposite  to  that  of  the  electron  and 
with  about  one-half  the  velocity,  since  the 
hole  mobility  is  about  one-half  the  electron 
mobility.  A sample  of  a semiconductor,  such  as 
germanium  or  silicon,  which  is  both  chemically 
pure  and  mechanically  perfect  will  contain  in  it 
approximately  equal  numbers  of  holes  and  elec- 
trons and  is  called  an  intrinsic  semiconductor. 
The  intrinsic  resistivity  of  the  semiconductor 
depends  strongly  upon  the  temperature,  being 
about  50  ohm/cm.  for  germanium  at  room 
temperature.  The  intrinsic  resistivity  of  silicon 
is  about  65,000  ohm/ cm.  at  the  same  temper- 
ature. 

If,  in  the  growing  of  the  semiconductor  crys- 
tal, a small  amount  of  an  impurity,  such  as 
phosphorous,  arsenic  or  antimony  is  included 
in  the  crystal,  each  atom  of  the  impurity  con- 
tributes one  free  electron.  This  electron  is 
available  for  conduction.  The  crystal  is  said 
to  be  doped  and  has  become  electron-conduct- 


Figure  2A 

CUT-AWAY  VIEW  OF  JUNCTION 
TRANSISTOR,  SHOWING  PHYSICAL 
ARRANGEMENT 


Figure  2B 

PICTORIAL  EQUIVALENT  OF 

P-N-P  JUNCTION  TRANSISTOR 


92  Transistors  and  Semi-Conductors 


THE  RADIO 


EMITTER.  .collector 


BASE- 
N-P-N  TRANSISTOR 


Figure  4 

ELECTRICAL  SYMBOLS 
FOR  TRANSISTORS 


Figure  3 

CONSTRUCTION  DETAIL  OF  A 
POINT  CONTACT  TRANSISTOR 


ing  in  nature  and  is  called  N (negative)  type 
germanium.  The  impurities  which  contribute 
electrons  are  called  donors.  N-type  germanium 
has  better  conductivity  than  pure  germanium  in 
one  direction,  and  a continuous  stream  of  elec- 
trons will  flow  through  the  crystal  in  this  direc- 
tion as  long  as  an  external  potential  of  the 
correct  polarity  is  applied  across  the  crystal. 

Other  impurities,  such  as  amminum,  gal- 
lium or  indium  add  one  hole  to  the  semicon- 
ducting crystal  by  accepting  one  electron  for 
each  atom  of  impurity,  thus  creating  additional 
holes  in  the  semiconducting  crystal.  The  ma- 
terial is  now  said  to  be  hole-conducting,  or  P 
(positive)  type  germanium.  The  impurities 
which  create  holes  are  called  acceptors.  P-type 
germanium  has  better  conductivity  than  pure 
germanium  in  one  direction.  This  direction  is 
opposite  to  that  of  the  N-type  material.  Either 
the  N-type  or  the  P-type  germanium  is  called 
extrinsic  conducting  type.  The  doped  materials 
have  lower  resistivities  than  the  pure  materials, 
and  doped  semiconductors  in  the  resistivity 
range  of  .01  to  10  ohm/cm.  are  normally  used 
in  the  production  of  transistors. 

5-3  The  Transistor 

In  the  past  few  years  an  entire  new  tech- 
nology has  been  developed  for  the  application 
of  certain  semiconducting  materials  in  produc- 
tion of  devices  having  gain  properties.  These 
gain  properties  were  previously  found  only  in 
vacuum  tubes.  The  elements  germanium  and 
silicon  are  the  principal  materials  which  ex- 
hibit the  proper  semiconducting  properties  per- 
mitting their  application  in  the  new  ampli- 
fying devices  called  transistors.  However, 
other  semiconducting  materials,  including  the 
compounds  indium  antimonide  and  lead  sulfide 
have  been  used  experimentally  in  the  produc- 
tion of  transistors. 


Types  of  Transistors  There  are  two  basic 
types  of  transistors,  the 
point-contact  type  and  the  junction  type  (fig- 
ure 2).  Typical  construction  detail  of  a point- 
contact  transistor  is  shown  in  Figure  3,  and 
the  electrical  symbol  is  shown  in  Figure  4.  The 
emitter  and  collector  electrodes  make  contact 
with  a small  block  of  germanium,  called  the 
base.  The  base  may  be  either  N-type  or  P-type 
germanium,  and  is  approximately  .05”  long 
and  .03”  thick.  The  emitter  and  collector  elec- 
trodes are  fine  wires,  and  are  spaced  about 
.005"  apart  on  the  germanium  base.  The  com- 
plete assembly  is  usually  encapsulated  in  a 
small,  plastic  case  to  provide  ruggedness  and 
to  avoid  contaminating  effects  of  the  atmos- 
phere. The  polarity  of  emitter  and  collector 
voltages  depends  upon  the  type  of  germanium 
employed  in  the  base,  as  illustrated  in  figure  4. 

The  junction  transistor  consists  of  a piece 
of  either  N-type  or  P-type  germanium  between 
two  wafers  of  germanium  of  the  opposite  type. 
Either  N-P-N  or  P-N-P  transistors  may  be 
made.  In  one  construction  called  the  grown 
crystal  process,  the  original  crystal,  grown 
from  molten  germanium  or  silicon,  is  created 
in  such  a way  as  to  have  the  two  closely  spaced 
junctions  imbedded  in  it.  In  the  other  con- 
struction called  the  fusion  process,  the  crystals 
are  grown  so  as  to  make  them  a single  con- 
ductivity type.  The  junctions  are  then  pro- 
duced by  fusing  small  pellets  of  special  metal 
alloys  into  minute  plates  cut  from  the  original 
crystal.  Typical  construction  detail  of  a junction 
transistor  is  shown  in  figure  2A. 

The  electrical  schematic  for  the  P-N-P  junc- 
tion transistor  is  the  same  as  for  the  point- 
contact  type,  as  is  shown  in  figure  4. 

Transistor  Action  Presently  available  types  of 
transistors  have  three  es- 
sential actions  which  collectively  are  called 
transistor  action.  These  are:  minority  carrier 
injection,  transport,  and  collection.  Figure  2B 
shows  a simplified  drawing  of  a P-N-P  junc- 
tion-type transistor,  which  can  illustrate  this 


HANDBOOK 


Transistors  93 


collective  action.  The  P-N-P  transistor  con- 
sists of  a piece  of  N-type  germanium  on  op- 
posite sides  of  which  a layer  of  P-type  mate- 
rial has  been  grown  by  the  fusion  process. 
Terminals  are  connected  to  the  two  P-sections 
and  to  the  N-type  base.  The  transistor  may  be 
considered  as  two  P-N  junction  rectifiers 
placed  in  close  juxaposition  with  a semi- 
conduction crystal  coupling  the  two  rectifiers 
together.  The  left-hand  terminal  is  biased  in 
the  forward  (or  conducting)  direction  and  is 
called  the  emitter.  The  right-hand  terminal  is 
biased  in  the  back  (or  reverse)  direction  and 
is  called  the  collector  The  operating  potentials 
are  chosen  with  respect  to  the  base  terminal, 
which  may  or  may  not  be  grounded.  If  an 
N-P-N  transistor  is  used  in  place  of  the  P-N-P, 
the  operating  potentials  are  reversed. 

The  Pe  - — Nb  junction  on  the  left  is  biased 
in  the  forward  direction  and  holes  from  the 
P region  are  injected  into  the  Nb  region,  pro- 
ducing therein  a concentration  of  holes  sub- 
stantially greater  than  normally  present  in  the 
material.  These  holes  travel  across  the  base 
region  towards  the  collector,  attracting  neigh- 
boring electrons,  finally  increasing  the  avail- 
able supply  of  conducting  electrons  in  the 
collector  loop.  As  a result,  the  collector  loop 
possesses  lower  resistance  whenever  the  emit- 
ter circuit  is  in  operation.  In  junction  tran- 
sistors this  charge  transport  is  by  means  of 
diffusion  wherein  the  charges  move  from  a 
region  of  high  concentration  to  a region  of 
lower  concentration  at  the  collector.  The  col- 
lector, biased  in  the  opposite  direction,  acts 
as  a sink  for  these  holes,  and  is  said  to  col- 
lect them. 

It  is  known  that  any  rectifier  biased  in  the 
forward  direction  has  a very  low  internal  im- 
pedance, whereas  one  biased  in  the  back  direc- 
tion has  a very  high  internal  impedance.  Thus, 
current  flows  into  the  transistor  in  a low  im- 
pedance circuit,  and  appears  at  the  output  as 
current  flowing  in  a high  impedance  circuit. 
The  ratio  of  a change  in  collector  current  to 
a change  in  emitter  current  is  called  the  current 
amplification,  or  alpha: 

Ic 

a — ~r- 

ie 

where  a = current  amplification 

ic  = change  in  collector  current 
ie  = change  in  emitter  current 

Values  of  alpha  up  to  3 or  so  may  be  ob- 
tained in  commercially  available  point-contact 
transistors,  and  values  of  alpha  up  to  about 
0.95  are  obtainable  in  junction  transistors. 


Alpha  Cutoff  The  alpha  cutoff  frequency  of 
Frequency  a transistor  is  that  frequency 

at  which  the  grounded  base 
current  gain  has  decreased  to  0.7  of  the  gain 
obtained  at  1 kc.  For  audio  transistors,  the 
alpha  cutoff  frequency  is  in  the  region  of  0.7 
Me.  to  1.5  Me.  For  r-f  and  switching  transis- 
tors, the  alpha  cutoff  frequency  may  be  5 Me. 
or  higher.  The  upper  frequency  limit  of  oper- 
ation of  the  transistor  is  determined  by  the 
small  but  finite  time  it  takes  the  majority  car- 
rier to  move  from  one  electrode  to  another. 

Drift  Transistors  As  previously  noted,  the 
signal  current  in  a con- 
ventional transistor  is  transmitted  across  the 
base  region  by  a diffusion  process.  The  transit 
time  of  the  carriers  across  this  region  is,  there- 
fore relatively  long.  RCA  has  developed  a 
technique  for  the  manufacture  of  transistors 
which  does  not  depend  upon  diffusion  for 
transmission  of  the  signal  across  the  base  re- 
gion. Transistors  featuring  this  new  process  are 
known  as  drift  transistors.  Diffusion  of  charge 
carriers  across  the  base  region  is  eliminated  and 
the  carriers  are  propelled  across  the  region  by 
a "built  in”  electric  field.  The  resulting  reduc- 
tion of  transit  time  of  the  carrier  permits  drift 
transistors  to  be  used  at  much  higher  fre- 
quencies than  transistors  of  conventional  de- 
sign. 

The  "built  in”  electric  field  is  in  the  base 
region  of  the  drift  transistor.  This  field  is 
achieved  by  utilizing  an  impurity  density 
which  varies  from  one  side  of  the  base  to  the 
other.  The  impurity  density  is  high  next  to 
the  emitter  and  low  next  to  the  collector.  Thus, 
there  are  more  mobile  electrons  in  the  region 
near  the  emitter  than  in  the  region  near  the 
collector,  and  they  will  try  to  diffuse  evenly 
throughout  the  base.  However,  any  displace- 
ment of  the  negative  charge  leaves  a positive 
charge  in  the  region  from  which  the  electrons 
came,  because  every  atom  of  the  base  material 
was  originally  electrically  neutral.  The  dis- 
placement of  the  charge  creates  an  electric 
field  that  tends  to  prevent  further  electron  dif- 
fusion so  that  a condition  of  equilibrium  is 
reached.  The  direction  of  this  field  is  such  as 
to  prevent  electron  diffusion  from  the  high 
density  area  near  the  emitter  to  the  low  density 
area  near  the  collector.  Therefore,  holes  enter- 
ing the  base  will  be  accelerated  from  the  emit- 
ter to  the  collector  by  the  electric  field.  Thus 
the  diffusion  of  charge  carriers  across  the  base 
region  is  augmented  by  the  built-in  electric 
field.  A potential  energy  diagram  for  a drift 
transistor  is  shown  in  figure  5. 


94  Transistors  and  Semi-Conductors 


THE  RADIO 


DECREASING 
POTENTIAL  ENERGY 

OP  MAJORITY  CARRIER 


Figure  5 

POTENTIAL  ENERGY  DIAGRAM 
FOR  DRIFT  TRANSISTOR  (2N247) 


5-4  Transistor 

Characteristics 

The  transistor  produces  results  that  may  be 
comparable  to  a vacuum  tube,  but  there  is  a 
basic  difference  between  the  two  devices.  The 
vacuum  tube  is  a voltage  controlled  device 
whereas  the  transistor  is  a current  controlled 
device.  A vacuum  tube  normally  operates  with 
its  grid  biased  in  the  negative  or  high  resist- 
ance direction,  and  its  plate  biased  in  the 
positive  or  low  resistance  direction.  The  tube 
conducts  only  by  means  of  electrons,  and  has 
its  conducting  counterpart  in  the  form  of  the 
N-P-N  transistor,  whose  majority  carriers  are 
also  electrons.  There  is  no  vacuum  tube  equiv- 
alent of  the  P-N-P  transistor,  whose  majority 
carriers  are  holes. 

The  biasing  conditions  stated  above  provide 
the  high  input  impedance  and  low  output  im- 
pedance of  the  vacuum  tube.  The  transistor  is 
biased  in  the  positive  or  low  resistance  direc- 
tion in  the  emitter  circuit,  and  in  the  negative, 
or  high  resistance  direction  in  the  collector 
circuit  resulting  in  a low  input  impedance 
and  a high  output  impedance,  contrary  to  and 
opposite  from  the  vacuum  tube.  A comparison 
of  point-contact  transistor  characteristics  and 
vacuum  tube  characteristics  is  made  in  figure  6. 

The  resistance  gain  of  a transistor  is  ex- 
pressed as  the  ratio  of  output  resistance  to 
input  resistance.  The  input  resistance  of  a 
typical  transistor  is  low,  in  the  neighborhood 
of  300  ohms,  while  the  output  resistance  is 
relatively  high,  usually  over  20,000  ohms.  For 
a point-contact  transistor,  the  resistance  gain 
is  usually  over  60. 

The  voltage  gain  of  a transistor  is  the 
product  of  alpha  times  the  resistance  gain, 
and  for  a point-contact  transistor  is  of  the 


COLLECTOR  VOLTS 
© 


PLATE  VOLTS 
® 

Figure  6 

COMPARISON  OF  POINT-CONTACT 
TRANSISTOR  AND  VACUUM  TUBE 
CHARACTERISTICS 


order  of  3 X 60  = 180.  A junction  transistor 
which  has  a value  of  alpha  less  than  unity 
nevertheless  has  a resistance  gain  of  the  order 
of  2000  because  of  its  extremely  high  output 
resistance,  and  the  resulting  voltage  gain  is 
about  1800  or  so.  For  both  types  of  transistors 
the  power  gain  is  the  product  of  alpha  squared 
times  the  resistance  gain  and  is  of  the  order 
of  400  to  500. 

The  output  characteristics  of  the  junction 
transistor  are  of  great  interest.  A typical  ex- 
ample is  shown  in  figure  7.  It  is  seen  that  the 
junction  transistor  has  the  characteristics  of 
an  ideal  pentode  vacuum  tube.  The  collector 
current  is  practically  independent  of  the  col- 
lector voltage.  The  range  of  linear  operation 
extends  from  a minimum  voltage  of  about  0.2 
volts  up  to  the  maximum  rated  collector  volt- 
age. A typical  load  line  is  shown,  which  il- 
lustrates the  very  high  load  impedance  that 
would  be  required  for  maximum  power  trans- 
fer. A grounded  emitter  circuit  is  usually  used, 
since  the  output  impedance  is  not  as  high  as 
when  a grounded  base  circuit  is  used. 


HANDBOOK 


Transistor  Characteristics  95 


Figure  7 

OUTPUT  CHARACTERISTICS  OF 
TYPICAL  JUNCTION  TRANSISTOR 


The  output  characteristics  of  a typical  point- 
contact  transistor  are  shown  in  figure  6.  The 
pentode  characteristics  ate  less  evident,  and  the 
output  impedance  is  much  lower,  with  the 
range  of  linear  operation  extending  down  to 
a collector  voltage  of  2 or  3.  Of  greater  prac- 
tical interest,  however,  is  the  input  charactet- 
istic  curve  with  short-circuited,  or  nearly  short- 
circuited  input,  as  shown  in  figure  8.  It  is 
this  point-contact  transistor  characteristic  of 
having  a region  of  negative  impedance  that 
lends  the  unit  to  use  in  switching  circuits.  The 
transistor  circuit  may  be  made  to  have  two, 
one  or  zero  stable  operating  points,  depending 
upon  the  bias  voltages  and  the  load  impedance 
used. 

Equivalent  Circuit  As  is  known  from  net- 

of  a Transistor  work  theory,  the  small 

signal  performance  of 
any  device  in  any  network  can  be  represented 
by  means  of  an  equivalent  circuit.  The  most 


Figure  8 

EMITTER  CHARACTERISTIC  CURVE 
FOR  TYPICAL  POINT  CONTACT 
TRANSISTOR 


cf  le 


VALUES  OF  THE  ECajIVALENT  CIRCUIT 


PARAMETER 

POINT- CONTACT 
TRANSISTOR 

Ue=  t MA.,  VC»15V.) 

JUNCTION 
TRANSISTOR  . 
(lE-  1 MA„  VC*  3V.) 

re  -EMITTER 
RESISTANCE 

100  A 

30  A 

rb-BASE 

RESISTANCE 

300  A 

300  A 

rc-  COLLECTOR 
RESISTANCE 

20000  A 

1 MEGOHM 

OC-CURRENT 

AMPLIFICATION 

2.0 

0.97 

Figure  9 

LOW  FREQUENCY  EQUIVALENT 
(Common  Bose)  CIRCUIT  FOR  POINT 
CONTACT  AND  JUNCTION 
TRANSISTOR 


convenient  equivalent  circuit  for  the  low  fre- 
quency small  signal  performance  of  both  point- 
contact  and  junction  transistors  is  shown  in 
figure  9.  Tf,  rb,  and  rc,  are  dynamic  resistances 
which  can  be  associated  with  the  emitter,  base 
and  collector  regions  of  the  transistor.  The 
current  generator  ah,  represents  the  transport 
of  charge  from  emitter  to  collector.  Typical 
values  of  the  equivalent  circuit  are  shown  in 
figure  9. 

Transistor  There  are  three  basic  transis- 

Configurotions  tor  configurations;  grounded 

base  connection,  grounded 
emitter  connection,  and  grounded  collector 
connection.  These  correspond  roughly  to 
grounded  grid,  grounded  cathode,  and  ground- 
ed plate  circuits  in  vacuum  tube  terminology 
(figure  10). 

The  grounded  base  circuit  has  a low  input 
impedance  and  high  output  impedance,  and  no 
phase  reversal  of  signal  from  input  to  output 
circuit.  The  grounded  emitter  circuit  has  a 
higher  input  impedance  and  a lower  output 
impedance  than  the  grounded  base  circuit,  and 
a reversal  of  phase  between  the  input  and  out- 
put signal  occurs.  This  circuit  usually  provides 
maximum  voltage  gain  from  a transistor.  The 
grounded  collector  circuit  has  relatively  high 
input  impedance,  low  output  impedance,  and 
no  phase  reversal  of  signal  from  input  to  out- 
put circuit.  Power  and  voltage  gain  are  both 

low. 

Figure  11  illustrates  some  practical  vacuum 
tube  circuits,  as  applied  to  transistors. 


96  Transistors  and  Semi-Conductors 


THE  RADIO 


i 


GROUNDED  BASE 
CONNECTION 


I ® 


GROUNDED  EMITTER 
CONNECTION 


GROUNDED  COLLECTOR 
CONNECTION 


Figure  10 

COMPARISON  OF  BASIC  VACUUM  TUBE  AND  TRANSISTOR  CONFIGURATIONS 


5-5  Transistor  Circuitry 

To  establish  the  correct  operating  parameters 
of  the  transistor,  a bias  voltage  must  be  estab- 
lished between  the  emitter  and  the  base.  Since 
transistors  are  temperature  sensitive  devices, 
and  since  some  variation  in  characteristics  usu- 
ally exists  between  transistors  of  a given  type, 
attention  must  be  given  to  the  bias  system  to 


overcome  these  difficulties.  The  simple  self-bias 
system  is  shown  in  figure  12 A.  The  base  is 
simply  connected  to  the  power  supply  through 
a large  resistance  which  supplies  a fixed  value 
of  base  current  to  the  transistor.  This  bias 
system  is  extremely  sensitive  to  the  current 
transfer  ratio  of  the  transistor,  and  must  be  ad- 
justed for  optimum  results  with  each  transistor. 

When  the  supply  voltage  is  fairly  high  and 


Figure  1 1 

TYPICAL  TRANSISTOR  CIRCUITS 


HANDBOOK 


Transistor  Circuitry  97 


-E 


-E 


© 


Figure  1 2 

BIAS  CONFIGURATIONS  FOR  TRANSISTORS. 


The  voltage  divider  system  of  C is  recommended  for  general  transistor  use.  Ratio  of  Ri/Rz 
establishes  base  bias,  and  emitter  bias  is  provided  by  voltage  drop  across  Re. 
Battery  Polarity  is  reversed  for  N-P-ff  transistors. 


wide  variations  in  ambient  temperature  do  not 
occur,  the  bias  system  of  figure  12B  may  be 
used,  with  the  bias  resistor  connected  from 
base  to  collector.  When  the  collector  voltage 
is  high,  the  base  current  is  increased,  moving 
the  operating  point  of  the  transistor  down  the 
load  line.  If  the  collector  voltage  is  low,  the 
operating  point  moves  upwards  along  the  load 
line,  thus  providing  automatic  control  of  the 
base  bias  voltage.  This  circuit  is  sensitive  to 
changes  in  ambient  temperature,  and  may  per- 
mit transistor  failure  when  the  transistor  is 
operated  near  maximum  dissipation  ratings. 

A better  bias  system  is  shown  in  figure  12C, 
where  the  base  bias  is  obtained  from  a voltage 
divider,  (Rl,  R2),  and  the  emitter  is  forward 
biased.  To  prevent  signal  degeneration,  the 
emitter  bias  resistor  is  bypassed  with  a large 
capacitance.  A high  degree  of  circuit  stability 
is  provided  by  this  form  of  bias,  providing  the 
emitter  capacitance  is  of  the  order  of  50  gfd. 
for  audio  frequency  applications. 

Audio  Circuitry  A simple  voltage  amplifier 
is  shown  in  figure  13.  Di- 
rect current  stabilization  is  employed  in  the 
emitter  circuit.  Operating  parameters  for  the 


amplifier  are  given  in  the  drawing.  In  this  case, 
the  input  impedance  of  the  amplifier  is  quite 
low.  When  used  with  a high  impedance  driv- 
ing source  such  as  a crystal  microphone  a step 
down  input  transformer  should  be  employed 
as  shown  in  figure  13B.  The  grounded  collec- 
tor circuit  of  figure  13C  provides  a high  input 
impedance  and  a low  output  impedance,  much 
as  in  the  manner  of  a vacuum  tube  cathode 
follower. 

The  circuit  of  a two  stage  resistance  coupled 
amplifier  is  shown  in  figure  14A.  The  input 
impedance  is  approximately  1100  ohms.  Feed- 
back may  be  placed  around  this  amplifier  from 
the  emitter  of  the  second  stage  to  the  base  of 
the  first  stage,  as  shown  in  figure  14B.  A 
direct  coupled  version  of  the  r-c  amplifier  is 
shown  in  figure  14C.  The  input  impedance  is 
of  the  order  of  15,000  ohms,  and  an  overall 
voltage  gain  of  80  may  be  obtained  with  a 
supply  potential  of  12  volts. 

It  is  possible  to  employ  N-P-N  and  P-N-P 
transistors  in  complementary  symmetry  circuits 
which  have  no  equivalent  in  vacuum  tube  de- 
sign. Figure  15A  illustrates  such  a circuit.  A 
symmetrical  push-pull  circuit  is  shown  in 


Figure  13 

P-N-P  TRANSISTOR  VOLTAGE  AMPLIFIERS 

A resistance  coupled  amplifier  employing  an  inexpensive  CK-722  transistor  is  shown  in  A.  For  use  with 
a high  impedance  crystal  microphone,  a step-down  transformer  matches  the  tow  input  impedance  of  the 
transistor,  as  shown  in  B.  The  grounded  collector  configuration  of  C provides  an  Input  impedance  of 

about  300,000  ohms. 


98  Transistors  and  Semi-Conductors 


THE  RADIO 


TWO  STAGE  TRANSISTOR  AUDIO  AMPLIFIERS 

The  feedback  loop  of  B may  be  added  to  the  r-c  amplifier  to  reduce  distortion,  or  to  control  the 
audio  response.  A direct  coupled  amplifier  is  shown  in  C. 


figure  15B.  This  circuit  may  be  used  to  di- 
rectly drive  a high  impedance  loudspeaker, 
eliminating  the  output  transformer.  A direct 
coupled  three  stage  amplifier  having  a gain 
figure  of  80  db  is  shown  in  figure  15C. 

The  transistor  may  also  be  used  as  a class 
A power  amplifier,  as  shown  in  figure  16A. 
Commercial  transistors  are  available  that  will 
provide  five  or  six  watts  of  audio  power  when 
operating  from  a 12  volt  supply.  The  smaller 
units  provide  power  levels  of  a few  milli- 
watts. The  correct  operating  point  is  chosen  so 
that  the  output  signal  can  swing  equally  in  the 
positive  and  negative  directions,  as  shown  in 
the  collector  curves  of  figure  16B. 

The  proper  primary  impedance  of  the  out- 
put transformer  depends  upon  the  amount  of 
power  to  be  delivered  to  the  load: 


The  collector  current  bias  is: 


In  a class  A output  stage,  the  maximum  a-c 


power  output  obtainable  is  limited  to  0.5  the 
allowable  dissipation  of  the  transistor.  The 
product  loEc  determines  the  maximum  collector 
dissipation,  and  a plot  of  these  values  is  shown 
in  figure  16B.  The  load  line  should  always  lie 
under  the  dissipation  curve,  and  should  encom- 
pass the  maximum  possible  area  between  the 
axes  of  the  graph  for  maximum  output  condi- 
tion. In  general,  the  load  line  is  tangent  to  the 
dissipation  curve  and  passes  through  the  supply 
voltage  point  at  zero  collector  current.  The  d-c 
operating  point  is  thus  approximately  one-half 
the  supply  voltage. 

The  circuit  of  a typical  push-pull  class  B 
transistor  amplifier  is  shown  in  figure  17A. 
Push-pull  operation  is  desirable  for  transistor 
operation,  since  the  even-order  harmonics  are 
largely  eliminated.  This  permits  transistors  to 
be  driven  into  high  collector  current  regions 
without  distortion  normally  caused  by  non- 
linearity of  the  collector.  Cross-over  distortion 
is  reduced  to  a minimum  by  providing  a slight 
forward  base  bias  in  addition  to  the  normal 
emitter  bias.  The  base  bias  is  usually  less  than 
0.5  volt  in  most  cases.  Excessive  base  bias  will 
boost  the  quiescent  collector  current  and  there- 
by lower  the  overall  efficiency  of  the  stage. 


Figure  15 

COMPLEMENTARY  SYMMETRY  AMPLIFIERS. 

N~P~N  and  P-N-P  transistors  may  be  combined  in  circuits  which  have  no  equivalent  in  vacuum  tube 
design.  Direct  coupling  between  cascaded  stages  using  o single  power  supply  source  may  be  employed,  as 
in  C.  Impedance  of  power  supply  should  be  extremely  low. 


HANDBOOK 


Transistor  Circuitry  99 


Figure  16 
TYPICAL  CLASS-A 
AUDIO  POWER 
TRANSISTOR  CIRCUIT. 

The  correct  operating  point  is 
chosen  so  that  output  signal  can 
swing  equally  in  a positive  or 
negative  direction,  without  ex- 
ceeding maximum  collector  dis- 
sipation. 


2N187A 


The  operating  point  of  the  class  B ampli- 
fier is  set  on  the  lc=0  axis  at  the  point  where 
the  collector  voltage  equals  the  supply  voltage. 
The  collector  to  collector  impedance  of  the 
output  transformer  is: 


In  the  class  B circuit,  the  maximum  a-c 
power  input  is  approximately  equal  to  five 
times  the  allowable  collector  dissipation  of 
each  transistor.  Power  transistors,  such  as  the 
2N301  have  collector  dissipation  ratings  of 
5.5  watts  and  operate  with  class  B efficiency 
of  about  67%.  To  achieve  this  level  of  opera- 
tion the  heavy  duty  transistor  relies  upon  ef- 
ficient heat  transfer  from  the  transistor  case 
to  the  chassis,  using  the  large  thermal  capacity 
of  the  chassis  as  a sink.  An  infinite  heat 
sink  may  be  approximated  by  mounting  the 
transistor  in  the  center  of  a 6"  x 6"  copper  or 
aluminum  sheet.  This  area  may  be  part  of  a 
larger  chassis. 

The  collector  of  most  power  transistors  is 
electrically  connected  to  the  case.  For  appli- 
cations where  the  collector  is  not  grounded  a 
thin  sheet  of  mica  may  be  used  between  the 
case  of  the  transistor  and  the  chassis. 

Power  transistors  such  as  the  Philco  T-1041 
may  be  used  in  the  common  collector  class  B 


configuration  (figure  17C)  to  obtain  high 
power  output  at  very  low  distortions  compar- 
able with  those  found  in  quality  vacuum  tube 
circuits  having  heavy  overall  feedback.  In  ad- 
dition, the  transistor  may  be  directly  bolted  to 
the  chassis,  assuming  a negative  grounded 
power  supply  Power  output  is  of  the  order  of 
10  watts,  with  about  0.5%  total  distortion. 

R-F  Circuitry  Transistors  may  be  used  for 
radio  frequency  work  provided 
the  alpha  cutoff  frequency  of  the  units  is 
sufficiently  higher  than  the  operating  fre- 
quency. Shown  in  figure  18A  is  a typical  i-f 
amplifier  employing  an  N-P-N  transistor.  The 
collector  current  is  determined  by  a voltage 
divider  on  the  base  circuit  and  by  a bias  re- 
sistor in  the  emitter  leg.  Input  and  output  are 
coupled  by  means  of  tuned  i-f  transformers. 
Bypass  capacitors  are  placed  across  the  bias 
resistors  to  prevent  signal  frequency  degener- 
ation. The  base  is  connected  to  a low  im- 
pedance untuned  winding  of  the  input  trans- 
former, and  the  collector  is  connected  to  a tap 
on  the  output  transformer  to  provide  proper 
matching,  and  also  to  make  the  performance  of 
the  stage  relatively  independent  of  variations 
between  transistors  of  the  same  type.  With  a 
rate-grown  N-P-N  transistor  such  as  the  G.E. 
2N293,  it  is  unnecessary  to  use  neutralization 
to  obtain  circuit  stability.  When  P-N-P  alloy 


® 


2N225  T-1041 


COLLECTOR  VOLTAGE  Ec 


Fisure  17 

CLASS-B  AUDIO  AMPLIFIER  CIRCUITRY. 


© 


The  common  collector  circuit  of  C permits  the  transistor  to  be  bolted  directly  to  the  chassis  for  efficient 
heat  transfer  from  the  transistor  case  to  the  chassis. 


100  Transistors  and  Semi-Conductors 


THE  RADIO 


TRANSISTORIZED  l-F  AMPLIFIERS. 

Typical  P-N-P  transhtor  must  be  neutralized  because  of  high  collector  capacitance.  Rate 
grown  N-P-N  transistor  does  not  usually  require  external  neutralizing  circuit. 


Figure  19 

AUTOMATIC  VOLUME 
CONTROL  CIRCUIT 
FOR  TRANSISTORIZED 
l-F  AMPLIFIER. 


transistors  are  used,  it  is  necessary  to  neutra- 
lize the  circuit  to  obtain  stability  (figure  18B). 

The  gain  of  a transistor  i-f  amplifier  will 
decrease  as  the  emitter  current  is  decreased. 
This  transistor  property  can  be  used  to  control 
the  gain  of  an  i-f  amplifier  so  that  weak  and 
strong  signals  will  produce  the  same  audio 
output.  A typical  i-f  strip  incorporating  this 
automatic  volume  control  action  is  shown  in 
figure  19. 

R-f  transistors  may  be  used  as  mixers  or 
autodyne  converters  much  in  the  same  manner 
as  vacuum  tubes  The  autodyne  circuit  is  shown 


COIL 


Figure  20 

THE  AUTODYNE  CONVERTER  CIRCUIT 
USING  A 2N168A  AS  A MIXER. 


signal  from  the  collector  to  the  emitter  caus- 
ing oscillation.  Capacitor  G tunes  the  oscillator 
circuit  to  a frequency  455  kc.  higher  than  that 
of  the  incoming  signal.  The  local  oscillator 
signal  is  inductively  coupled  into  the  emitter 
circuit  of  the  transistor.  The  incoming  signal 
is  resonated  in  Ts  and  coupled  via  a low  im- 
pedance winding  to  the  base  circuit.  Notice 
that  the  base  is  biased  by  a voltage  divider 
circuit  much  the  same  as  is  used  in  audio  fre- 
quency operation.  The  two  signals  are  mixed 
in  this  stage  and  the  desired  beat  frequency  of 
455  kc.  is  selected  by  i-f  transformer  Ts  and 
passed  to  the  next  stage.  Collector  currents  of 
0.6  ma.  to  0.8  ma.  are  common,  and  the  local 
oscillator  injection  voltage  at  the  emitter  is  in 
the  range  of  0.15  to  0.25  volts,  r.m.s. 

A complete  receiver  "front  end”  capable  of 
operation  up  to  23  Me.  is  shown  in  figure  21. 
The  RCA  2N247  drift  transistor  is  used  for 
the  r-f  amplifier  (TRl),  mixer  (TR2),  and 
high  frequency  oscillator  (TR3).  The  2N247 
incorporates  an  interlead  shield,  cutting  the 
interlead  capacitance  to  .003  fi/iid.  If  proper 
shielding  is  employed  between  the  tuned  cir- 
cuits of  the  r-f  stage,  no  neutralization  of  the 
stage  is  required.  The  complete  assembly  ob- 
tains power  from  a 9-volt  transistor  battery. 
Note  that  input  and  output  circuits  of  the  tran- 
sistors are  tapped  at  low  impedance  points  on 
the  r-f  coils  to  achieve  proper  impedance  match. 


HANDBOOK 


Transistor  Circuitry  101 


Figure  21 

RF  AMPLIFIER,  MIXER, 
AND  OSCILLATOR 
STAGES  FOR 
TRANSISTORIZED 
HIGH  FREQUENCY 
RECEIVER.  THE  RCA 
2N247  DRIFT 
TRANSISTOR  IS 
CAPABLE  OF 
EFFICIENT  OPERATION 
UP  TO  23  Me. 


Transistor  Sufficient  coupling  of  the  proper 
Oscillators  phase  between  input  and  output 
circuits  of  the  transistor  will  per- 
mit oscillation  up  to  and  slightly  above  the 
alpha  cutoff  frequency.  Various  forms  of  tran- 
sistor oscillators  are  shown  in  figure  22.  A 
simple  grounded  emitter  Hartley  oscillator  hav- 
ing positive  feedback  between  the  base  and  the 
collector  (22A)  is  compared  to  a grounded 
base  Hartley  oscillator  (22B).  In  each  case 
the  resonant  tank  circuit  is  common  to  the  in- 
put and  output  circuits  of  the  transistor.  Self- 
bias of  the  transistor  is  employed  in  both  these 
circuits  A mote  sophisticated  oscillator  em- 
ploying a 2N247  transistor  and  utilizing  a 
voltage  divider-type  bias  system  (figure  22C) 
is  capable  of  operation  up  to  50  Me.  or  so. 
The  tuned  circuit  is  placed  in  the  collector, 
with  a small  emitter-collector  capacitor  provid- 
ing feedback  to  the  emitter  electrode. 

A P-N-P  and  an  N-P-N  transistor  may  be 
combined  to  form  a complementary  Hartley 
oscillator  of  high  stability  (figure  23)-  The 
collector  of  the  P-N-P  transistor  is  directly 


coupled  to  the  base  of  the  N-P-N  transistor, 
and  the  emitter  of  the  N-P-N  transistor  furn- 
ishes the  correct  phase  reversal  to  sustain  os- 
cillation. Heavy  feedback  is  maintained  be- 
tween the  emitter  of  the  P-N-P  transistor  and 
the  collector  of  N-P-N  transistor.  The  degree 
of  feedback  is  controlled  by  Ri.  The  emitter 
resistor  of  the  second  transistor  is  placed  at  the 


Figure  23 

COMPLEMENTARY  HARTLEY 
OSCILLATOR 


P-N-P  and  N-P~N  transistors  form  high  sta- 
bility oscillator.  Feedback  between  P-N-P 
emitter  and  N-P-N  collector  is  controlled  by 
Pi.  1NB1  diodes  are  used  as  amplitude  lim- 
iters. Frequency  of  oscillation  is  determined 
by  L,  Ci-Ct. 


Figure  22 

TYPICAL  TRANSISTOR  OSCILLATOR  CIRCUITS 

A — Grounded  Emitter  Hartley 
B — Grounded  Base  Hartley 

C — 2N247  Oscillator  Suitable  for  SO  Me.  operation. 


102  Transistors  and  Semi-Conductors 


THE  RADIO 


2N33 

POINT-CONTACT  TRANSISTOR 


NEGATIVE  RESISTANCE  OF 
POINT-CONTACT  TRANSISTOR 
PERMITS  HIGH  FREQUENCY 
OSCILLATION  (50  Me)  WITHOUT 
WITHOUT  NECESSITY  OF 
EXTERNAL  FEEDBACK  PATH. 


center  of  the  oscillator  coil  to  eliminate  load- 
ing of  the  tuned  circuit. 

Two  germanium  diodes  are  employed  as 
amplitude  limiters,  further  stabilizing  ampli- 
fier operation.  Because  of  the  low  circuit  im- 
pedances, it  is  permissible  to  use  extremely 
high-C  in  the  oscillator  tank  circuit,  effectively 
limiting  oscillator  temperature  stability  to  var- 
iations in  the  tank  inductance. 

The  point-contact  transistor  exhibits  nega- 
tive input  and  output  resistances  over  part  of 
its  operaing  range,  due  to  its  unique  ability 
to  multiply  the  input  current.  This  character- 
istic affords  the  use  of  oscillator  circuitry  hav- 
ing no  external  feedback  paths  (figure  24). 
A high  impedance  resonant  circuit  in  the  base 
lead  produces  circuit  instability  and  oscillation 
at  the  resonant  frequency  of  the  L-C  circuit. 
Positive  emitter  bias  is  used  to  insure  thermal 
circuit  stability. 


POINT-CONTACT 

TRANSISTOR 


Figure  25 

RELAXATION  OSCILLATOR  USING 
POINT-CONTACT  OR  SURFACE 
BARRIER  TRANSISTORS. 


Relaxation  Transistors  have  almost  unlimit- 
Oscillators  ed  use  in  relaxation  and  R-C  os- 
cillator service.  The  negative  re- 
sistance characteristic  of  the  point  contact  tran- 
sistor make  it  well  suited  to  such  application. 
Surface  barrier  transistors  are  also  widely  used 
in  this  service,  as  they  have  the  highest  alpha 
cutoff  frequency  among  the  group  of  "alpha- 
less-than-unity”  transistors.  Relaxation  oscilla- 
tors used  for  high  speed  counting  require  tran- 
sistors capable  of  operation  at  repetition  rates 
of  5 Me.  to  10  Me. 

A simple  emitter  controlled  relaxation  os- 
cillator is  shown  in  figure  25,  together  with 
its  operating  characteristic.  The  emitter  of  the 
transistor  is  biased  to  cutoff  at  the  start  of  the 
cycle  (point  1 ).  The  charge  on  the  emitter  ca- 
pacitor slowly  leaks  to  ground  through  the 
emitter  resistor,  Ri.  Discharge  time  is  deter- 
mined by  the  time  constant  of  RiG.  When  the 
emitter  voltage  drops  sufficiently  low  to  permit 
the  transistor  to  reach  the  negative  resistance 
region  (point  2)  the  emitter  and  collector  re- 
sistances drop  to  a low  value,  and  the  collector 


Figure  26 

TRANSISTORIZED  BLOCKING  OSCILLATOR  (A)  AND  ECCLES-JORDAN 
BI-STABLE  MULTIVIBRATOR  (B). 

High-alpha  transistors  must  be  employed  in  counting  circuits  to  reduce  effects  of 
storage  time  caused  by  transit  lag  in  transistor  base. 


HANDBOOK 


Transistor  Circuitry  103 


Figure  28 

SCHEMATIC,  TRANSISTORIZED  BROADCAST  BAND  (500-  1600  KC.)  SUPERHETERO- 
DYNE RECEIVER. 


Figure  27 

"WRIST  RADIO"  CAN  BE  MADE 
WITH  LOOPSTICK,  DIODE,  AND 
INEXPENSIVE  CK-722  TRANSISTOR. 

A TWENTY  FOOT  ANTENNA  WIRE 
WILL  PROVIDE  GOOD  RECEPTION 
IN  STRONG  SIGNAL  AREAS. 

current  is  limited  only  by  the  collector  resistor, 
R2.  The  collector  current  is  abruptly  reduced 
by  the  charging  action  of  the  emitter  capacitor 
Cl  (point  3),  bringing  the  circuit  back  to  the 
original  operating  point.  The  "spike”  of  col- 
lector current  is  produced  during  the  charging 
period  of  Ci.  The  duration  of  the  pulse  and  the 
pulse  repetition  frequency  (p.r.f. ) are  con- 
trolled by  the  values  of  Ci,  Ri,  R:,  and  Ri. 

Transistors  may  also  be  used  as  blocking 
oscillators  (figure  26A).  The  oscillator  may 
be  synchronized  by  coupling  the  locking  signal 
to  the  base  circuit  of  the  transistor.  An  oscil- 
lator of  this  type  may  be  used  to  drive  a flip- 
flop  circuit  as  a counter.  An  Eccles-Jordan 
bi-stable  flip-flop  circuit  employing  surface- 
barrier  transistors  may  be  driven  between  "off" 
and  "on”  positions  by  an  exciting  pulse  as 
shown  in  figure  26B.  The  first  pulse  drives 
the  "on”  transistor  into  saturation.  This  tran- 
sistor remains  in  a highly  conductive  state  until 
the  second  exciting  pulse  arrives.  The  transis- 
tor does  not  immediately  return  to  the  cut-off 


state,  since  a time  lapse  occurs  before  the  out- 
put waveform  starts  to  decrease.  This  storage 
time  is  caused  by  the  transit  lag  of  the  minority 
carriers  in  the  base  of  the  transistor.  Proper  cir- 
cuit design  and  the  use  of  high-alpha  transis- 
tors can  reduce  the  effects  of  storage  time  to  a 
minimum.  Driving  pulses  may  be  coupled  to 
the  multivibrator  through  steering  diodes  as 
shown  in  the  illustration. 

5-6  Transistor  Circuits 

With  the  introduction  of  the  dollar  tran- 
sistor, many  interesting  and  unusual  experi- 
ments and  circuits  may  be  built  up  by  the  be- 
ginner in  the  transistor  field.  One  of  the  most 
interesting  is  the  "wrist  watch”  receiver,  illus- 
trated in  figure  27.  A diode  and  a transistor 
amplifier  form  a miniature  broadcast  receiver, 
which  may  be  built  in  a small  box  and  carried 
on  the  person.  A single  1.5-volt  penlite  cell 
provides  power  for  the  transistor,  and  a short 
length  of  antenna  wire  will  suffice  in  the  vi- 
cinity of  a local  broadcasting  station. 

A transistorized  superhetrodyne  for  broad- 
cast reception  is  shown  in  figure  28.  No  an- 
tenna is  required,  as  a ferrite  "loop-stick”  is 
used  for  the  r-f  input  circuit  of  the  2N136 
mixer  transistor.  A miniature  magnetic  "hear- 
ing aid”  type  earphone  may  be  employed  with 
this  receiver. 

A simple  phonograph  amplifier  designed 
for  use  with  a high  impedance  crystal  pickup 
is  shown  in  figure  29.  Two  stages  of  amplifi- 
cation using  2N109  transistors  are  used  to 
drive  two  2N109  transistors  in  a class  B con- 
figuration. Approximately  200  milliwatts  of 


104  Transistors  and  Semi-Conductors 


Figure  29 

HIGH  GAIN,  LOW  DISTORTION  AUDIO  AMPLIFIER,  SUITABLE  FOR  USE 
WITH  A CRYSTAL  PICKUP.  POWER  OUTPUT  IS  250  MILLIWATTS. 


power  may  be  obtained  with  a battery  supply  esting  transistor  projects  will  be  shown  in  later 

of  12  volts.  Peak  current  drain  under  maxi-  chapters  of  this  Handbook, 

mum  signal  conditions  is  40  ma.  Other  inter- 


E & E TECHNI-SHEET 

TABLE  OF  WROUGHT  STEEL  STANDARD 
{BLACK  OR  GALVANIZED)  RANDOM  LENGTHS, 

"nOmTn SIZE  'oTo  LD^ 

INCHES INCHES INCHES 

1/8" .405 .269 

1/4" .540 .364 

3/8" .675 .493 

1/2" .840 .622 

3/4" 1.050 .824 

1 .31  5 1 .04  9 

1 1 /4  " 1 .660 1 .380 

1 1/2" 1 .900 1.610 

^ 2.375 2 .067 

2 1/2" 2.875 2.469 

3"  3.500  3.068 


3 1/2"  4.000  3.548 


4"  4.500  4.026 


PI  PE 

20-21  FEET 

poi7nds~ 

PER  FOOT 
.244 
■ 424 

^6_7 

.850 

1.130 

1 .678 
2.272 

2 .71  7 
3.652 
5 .793 
7.575 


9 .109 


10.790 


TABLE  OF  ELECTRICAL  METALLIC  TUBING  (EMT) 
GALVANIZED,  10  FOOT  LENGTHS 


NOMINAL  S 

INCHES 

3/8" 

1/2" 

3/4" 



1 1 /4  " 

1 1 /2  " 

2" 


O.D. 

INCHES 

l.D. 

INCHES 

POUNDS 

PER  FOOT 

.577 

.493 

.250 

.706 

.622 

.321 

.922 

.824 

.488 

1.16  3 

1 .049 

.711 

1.508 

1.308 

1 .000 

1 .738 

1.610 

1.1  80 

2.195 

2.067 

1.500 

TABLE  OF  ALUMINUM  ROD,  12  FOOT  LENGTHS 

DIAMETER  WT.  */FT. 

1/8" .01  5 

3/16" 

1 /4  " 


CHAPTER  SIX 


Vacuum  Tube  Amplifiers 


6-1  Vacuum  Tube  Parameters 

The  ability  of  the  control  grid  of  a vacuum 
tube  to  control  large  amounts  of  plate  power 
with  a small  amount  of  grid  energy  allows  the 
vacuum  tube  to  be  used  as  an  amplifier.  It  is 
this  ability  of  vacuum  tubes  to  amplify  an 
extremely  small  amount  of  energy  up  to  almost 
any  level  without  change  in  anything  except 
amplitude  which  makes  the  vacuum  tube  such 
an  extremely  valuable  adjunct  to  modern  elec- 
tronics and  communication. 

Symbols  for  As  an  assistance  in  simplify- 
Vocuum-Tube  ing  and  shortening  expressions 
Parameters  involving  vacuum-tube  param- 
eters, the  following  symbols 
will  be  used  throughout  this  book: 

Tube  Constants 

fi  — amplification  factor 
Rp  — plate  resistance 
Gm  — transconductance 

— grid-screen  mu  factor 
Gc  — conversion  transconductance(mixer  tube) 

Interelectrode  Capacitances 

Cgi;  — grid-cathode  capacitance 

Cgp  — grid-plate  capacitance 

Cpii  — plate-cathode  capacitance 

Ci„ — input  capacitance  (tetrode  or  pentode) 

Cpu,  — output  capacitance  (tetrode  or  pentode) 


Electrode  Potentials 

Ebb — d-c  plate  supply  voltage  (a  positive 
quantity) 

Epc — d-c  grid  supply  voltage  (a  negative 
quantity) 

Egp, — peak  grid  excitation  voltage  ()^  total 
peak-to-peak  grid  swing) 

Ep,„  — peak  plate  voltage  (/i  total  peak-to-peak 
plate  swing) 

ep  — instantaneous  plate  potential 
eg  — instantaneous  grid  potential 
^pmin  — minimum  instantaneous  plate  voltage 
egmp  — maximum  positive  instantaneous  grid 
voltage 

Ep  — static  plate  voltage 
Eg  — static  grid  voltage 
Cpo  — cutoff  bias 

Electrode  Currents 

Ib  — average  plate  current 

Ip  — average  grid  current 

1pm  — peak  fundamental  plate  current 

Ipmax  — maximum  instantaneous  plate  current 

Igraax  — maximum  instantaneous  grid  current 

Ip  — static  plate  current 

Ig  — static  grid  current 

Other  Symbols 

P;  — plate  power  input 
P„  — plate  power  output 
Pp  — plate  dissipation 

Pj  — grid  driving  power  (grid  plus  bias  losses) 


106 


Classes  of  Amplifiers  107 


TRIODE  PENTODE  OR  TETRODE 


Figure  1 

STATIC  INTERELECTRODE  CAPACI- 
TANCES WITHIN  A TRIODE,  PENTODE, 
OR  TETRODE 


Pg  — grid  dissipation 

Np  — plate  efficiency(exptessed  as  a decimal) 
0p  — one-half  angle  of  plate  current  flow 
0g  — one-half  angle  of  grid  current  flow 
Rl  — load  resistance 
Zl  — load  impedance 

Vacuum-Tube  The  relationships  between  cer- 
Constants  tain  of  the  electrode  potentials 
and  currents  within  a vacuum 
tube  are  reasonably  constant  under  specified 
conditions  of  operation.  These  relationships 
are  called  vacuum-tube  constants  and  are 
listed  in  the  data  published  by  the  manufac- 
turers of  vacuum  tubes.  The  defining  equations 
for  the  basic  vacuum-tube  constants  are  given 
in  Chapter  Four. 


Interelectrode  The  values  of  interelectrode 
Capacitances  and  capacitance  published  in 
Miller  Effect  vacuum-tube  tables  are  the 

static  values  measured,  in 
the  case  of  triodes  for  example,  as  shown  in 
figure  1.  The  static  capacitances  ate  simply 
as  shown  in  the  drawing,  but  when  a tube  is 
operating  as  amplifier  there  is  another  con- 
sideration known  as  Miller  Effect  which  causes 
the  dynamic  input  capacitance  to  be  different 
from  the  static  value.  The  output  capacitance 
of  an  amplifier  is  essentially  the  same  as  the 
static  value  given  in  the  published  tube  tables. 
The  grid-to-plate  capacitance  is  also  the  same 
as  the  published  static  value,  but  since  the 
Cgp  acts  as  a small  capacitance  coupling  en- 
ergy back  from  the  plate  circuit  to  the  grid 
circuit,  the  dynamic  input  capacitance  is  equal 
to  the  static  value  plus  an  amount  (frequently 
much  greater  in  the  case  of  a triode)  deter- 
mined by  the  gain  of  the  stage,  the  plate  load 
impedance,  and  the  Cgp  feedback  capacitance. 
The  total  value  for  an  audio  amplifier  stage 
can  be  expressed  in  the  following  equation: 

(^(dynamic)  ^ ^(static) 
where  is  the  grid-to-cathode  capacitance. 


Cgp  is  the  grid-to-plate  capacitance,  and  A is 
the  stage  gain.  This  expression  assumes  that 
the  vacuum  tube  is  operating  into  a resistive 
load  such  as  would  be  the  case  with  an  audio 
stage  working  into  a resistance  plate  load  in 
the  middle  audio  range. 

The  more  complete  exptession  for  the  input 
admittance  (vectot  sum  of  capacitance  and 
resistance)  of  an  amplifier  operating  into  any 
type  of  plate  load  is  as  follows: 

Input  capacitance  = Cgjj  -t  (1  -I-  A cos  d)  Cgp 


A sin  d 

Where:  Cg^  = grid-to-cathode  capacitance 
Cgp  = grid-to-plate  capacitance 
A = voltage  amplification  of  the  tube 
alone 

6 - phase  angle  of  the  plate  load  im- 

pedance, positive  fot  inductive 
loads,  negative  fot  capacitive 

It  can  be  seen  from  the  above  that  if  the 
plate  load  impedance  of  the  stage  is  capaci- 
tive or  inductive,  there  will  be  a resistive  com- 
ponent in  the  input  admittance  of  the  stage. 
The  resistive  component  of  the  input  admit- 
tance will  be  positive  (tending  to  load  the 
circuit  feeding  the  grid)  if  the  load  impedance 
of  the  plate  is  capacitive,  or  it  will  be  negative 
(tending  to  make  the  stage  oscillate)  if  the 
load  impedance  of  the  plate  is  inductive. 

Neutralization  Neutralization  of  the  effects 

of  Interelectrode  of  interelectrode  capacitance 

Capacitance  is  employed  most  frequently 

in  the  case  of  radio  fre- 
quency power  amplifiers.  Before  the  introduc- 
tion of  the  tetrode  and  pentode  tube,  triodes 
were  employed  as  neutralized  Class  A ampli- 
fiers in  receivers.  This  practice  has  been 
largely  superseded  in  the  present  state  of  the 
art  through  the  use  of  tetrode  and  pentode 
tubes  in  which  the  Cgp  or  feedback  capaci- 
tance has  been  reduced  to  such  a low  value 
that  neutralization  of  its  effects  is  not  neces- 
sary to  prevent  oscillation  and  instability. 


6-2  Classes  and  Types  of 
Vacuum-Tube  Amplifiers 

Vacuum-tube  amplifiers  are  grouped  into 
various  classes  and  sub-classes  according  to 
the  type  of  work  they  are  intended  to  perfotm. 
The  difference  between  the  various  classes  is 
determined  primarily  by  the  value  of  average 
grid  bias  employed  and  the  maximum  value  of 


108  Vacuum  Tube  Amplifiers 


THE  RADIO 


the  exciting  signal  to  be  impressed  upon  the 

grid. 

Class  A A Class  A amplifier  is  an  amplifier 
Amplifier  biased  and  supplied  with  excitation 
of  such  amplitude  that  plate  cur- 
rent flows  continuously  (360°  of  the  exciting 
voltage  waveshape)  and  grid  current  does  not 
flow  at  any  time.  Such  an  amplifier  is  normally 
operated  in  the  center  of  the  grid-voltage 
plate-current  transfer  characteristic  and  gives 
an  output  waveshape  which  is  a substantial 
replica  of  the  input  waveshape. 

Class  Aj  This  is  another  term  applied  to  the 
Amplifier  Class  A amplifier  in  which  grid 
current  does  not  flow  over  any 
portion  of  the  input  wave  cycle. 

Class  A2  This  is  a Class  A amplifier  oper- 
Amplifier  ated  under  such  conditions  that  the 
grid  is  driven  positive  over  a por- 
tion of  the  input  voltage  cycle,  but  plate  cur- 
rent still  flows  over  the  entire  cycle. 

Class  ABj  This  is  an  amplifier  operated  under 
Amplifier  such  conditions  of  grid  bias  and 
exciting  voltage  that  plate  current 
flows  for  more  than  one-half  the  input  voltage 
cycle  but  for  less  than  the  complete  cycle.  In 
other  words  the  operating  angle  of  plate  cur- 
rent flow  is  appreciably  greater  than  180°  but 
less  than  360°.  The  suffix  1 indicates  that  grid 
current  does  not  flow  over  any  portion  of  the 
input  cycle. 

Class  AB2  A Class  AB2  amplifier  is  operated 
Amplifier  under  essentially  the  same  condi- 
tions of  grid  bias  as  the  Class 
AB  j amplifier  mentioned  above,  but  the  excit- 
ing voltage  is  of  such  amplitude  that  grid  cur- 
rent flows  over  an  appreciable  portion  of  the 
input  wave  cycle. 

Class  B A Class  B amplifier  is  biased  sub- 
Amplifier  stantially  to  cutoff  of  plate  current 
(without  exciting  voltage)  so  that 
plate  current  flows  essentially  over  one-half 
the  input  voltage  cycle.  The  operating  angle 
of  plate  current  flow  is  essentially  180°.  The 
Class  B amplifier  is  almost  always  excited 
to  such  an  extent  that  grid  current  flows. 

Class  C A Class  C amplifier  is  biased  to  a 
Amplifier  value  greater  than  the  value  re- 
quired for  plate  current  cutoff  and 
is  excited  with  a signal  of  such  amplitude 
that  grid  current  flows  over  an  appreciable 
period  of  the  input  voltage  waveshape.  The 
angle  of  plate  current  flow  in  a Class  C am- 
plifier is  appreciably  less  than  180°,  or  in 
other  words,  plate  current  flows  appreciably 


Figure  2 

TYPES  OF  BIAS  SYSTEMS 

A - Grid  bias 
B - Cathode  bias 
C - Grid  leak  bias 


less  than  one-half  the  time.  Actually,  the  con- 
ventional operating  conditions  for  a Class  C 
amplifier  are  such  that  plate  current  flows  for 
120°  to  150°  of  the  exciting  voltage  wave- 
shape. 

Types  of  There  are  three  general  types  of 
Amplifiers  amplifier  circuits  in  use.  These 
types  are  classified  on  the  basis 
of  the  return  for  the  input  and  output  circuits. 
Conventional  amplifiers  are  called  cathode  re- 
turn amplifiers  since  the  cathode  is  effectively 
grounded  and  acts  as  the  common  return  for 
both  the  input  and  output  circuits.  The  second 
type  is  known  as  a plate  return  amplifier  or 
cathode  follower  since  the  plate  circuit  is  ef- 
fectively at  ground  for  the  input  and  output 
signal  voltages  and  the  output  voltage  or 
power  is  taken  between  cathode  and  plate.  The 
third  type  is  called  a grid-return  or  grounded- 
grid  amplifier  since  the  grid  is  effectively  at 
ground  potential  for  input  and  output  signals 
and  output  is  taken  between  grid  and  plate. 


6-3  Biasing  Methods 

The  difference  of  potential  between  grid  and 
cathode  is  called  the  grid  bids  of  a vacuum 
tube.  There  are  three  general  methods  of 
providing  this  bias  voltage.  In  each  of  these 
methods  the  purpose  is  to  establish  the  grid 
at  a potential  with  respect  to  the  cathode 
which  will  place  the  tube  in  the  desired  opera- 
ting condition  as  determined  by  its  charac- 
teristics. 

Grid  bias  may  be  obtained  from  a source  of 
voltage  especially  provided  for  this  purpose, 
as  a battery  or  other  d-c  power  supply.  This 
method  is  illustrated  in  figure  2A,  and  is 
known  as  fixed  bias. 

A second  biasing  method  is  illustrated  in 
figure  2B  which  utilizes  a cathode  resistor 
across  which  an  IR  drop  is  developed  as  a 
result  of  plate  current  flowing  through  it.  The 


HANDBOOK 


Amplifier  Distortion  109 


cathode  of  the  tube  is  held  at  a positive  po- 
tential with  respect  to  ground  by  the  amount  of 
the  IR  drop  because  the  grid  is  at  ground  po- 
tential. Since  the  biasing  voltage  depends 
upon  the  flow  of  plate  current  the  tube  cannot 
be  held  in  a cutoff  condition  by  means  of  the 
cathode  bias  voltage  developed  across  the 
cathode  resistor.  The  value  of  this  resistor  is 
determined  by  the  bias  required  and  the  plate 
current  which  flows  at  this  value  of  bias,  as 
found  from  the  tube  characteristic  curves. 
A capacitor  is  shunted  across  the  bias  resistor 
to  provide  a low  impedance  path  to  ground  for 
the  a-c  component  of  the  plate  current  which 
results  from  an  a-c  input  signal  on  the  grid. 

The  third  method  of  providing  a biasing 
voltage  is  shown  in  figure  2C,  and  is  called 
grid-leak  bias.  During  the  portion  of  the  input 
cycle  which  causes  the  grid  to  be  positive 
with  respect  to  the  cathode,  grid  current  flows 
from  cathode  to  grid,  charging  capacitor  Cg. 
When  the  grid  draws  current,  the  grid-to-cathoae 
resistance  of  the  tube  drops  from  an  infinite 
value  to  a very  low  value,  on  the  order  of 

1.000  ohms  or  so,  making  the  charging  time 
constant  of  the  capacitor  very  short.  This  en- 
ables Cg  to  charge  up  to  essentially  the  full 
value  of  the  positive  input  voltage  and  results 
in  the  grid  (which  is  connected  to  the  low  po- 
tential plate  of  the  capacitor)  being  held  es- 
sentially at  ground  potential.  During  the  nega- 
tive swing  of  the  input  signal  no  grid  current 
flows  and  the  discharge  path  of  Cg  is  through 
the  grid  resistance  which  has  a value  of 

500.000  ohms  or  so.  The  discharge  time  con- 
stant for  Cg  is,  therefore,  very  long  in  com- 
parison to  the  period  of  the  input  signal  and 
only  a small  part  of  the  charge  on  Cg  is  lost. 
Thus,  the  bias  voltage  developed  by  the  dis- 
charge of  Cg  is  substantially  constant  and  the 
grid  is  not  permitted  to  follow  the  positive 
portions  of  the  input  signal. 

6-4  Distortion  in  Amplifiers 

There  are  three  main  types  of  distortion  that 
may  occur  in  amplifiers;  frequency  distortion, 
phase  distortion  and  amplitude  distortion. 

Frequency  Frequency  distortion  may  occur 
Distortion  when  some  frequency  components 
of  a signal  ate  amplified  more  than 
others.  Frequency  distortion  occurs  at  low 
frequencies  if  coupling  capacitors  between 
stages  are  too  small,  or  may  occur  at  high  fre- 
quencies as  a result  of  the  shunting  effects  of 
the  distributed  capacities  in  the  circuit. 

Phase  In  figure  3 an  input  signal  con- 
Distortioh  sisting  of  a fundamental  and  a 
third  harmonic  is  passed  through 


Figure  3 

lllustralion  of  the  effect  of  phase  distortion 
on  input  wave  containing  a third  harmonic 
signal 


a two  stage  amplifier.  Although  the  amplitudes 
of  both  components  are  amplified  by  identical 
ratios,  the  output  waveshape  is  considerably 
different  from  the  input  signal  because  the 
phase  of  the  third  harmonic  signal  has  been 
shifted  with  respect  to  the  fundamental  signal. 
This  phase  shift  is  known  as  phase  distortion, 
and  is  caused  principally  by  the  coupling  cir- 
cuits between  the  stages  of  the  amplifier. 
Most  coupling  circuits  shift  the  phase  of  a 
sine  wave,  but  this  has  no  effect  on  the  shape 
of  the  output  wave.  However,  when  a complex 
wave  is  passed  through  the  same  coupling 
circuit,  each  component  frequency  of  the  wave- 
shape may  be  shifted  in  phase  by  a different 
amount  so  that  the  output  wave  is  not  a faith- 
ful reproduction  of  the  input  waveshape. 

Amplitude  If  a signal  is  passed  through  a vac- 
Distortion  uum  tube  that  is  operating  on  any 
non-linear  part  of  its  characteristic, 
amplitude  distortion  will  occur.  In  such  a re- 
gion, a change  in  grid  voltage  does  not  result 
in  a change  in  plate  current  which  is  directly 
proportional  to  the  change  in  grid  voltage.  For 
example,  if  an  amplifier  is  excited  with  a sig- 
nal that  overdrives  the  tubes,  the  resultant 
signal  is  distorted  in  amplitude,  since  the 
tubes  operate  over  a non-linear  portion  of  their 
characteristic. 


6-5  Resistance- 

Capacitance  Coupled 
Audio-Frequency  Amplifiers 

Present  practice  in  the  design  of  audio-fre- 
quency voltage  amplifiers  is  almost  exclusively 
to  use  resistance-capacitance  coupling  be- 
tween the  low-level  stages.  Both  triodes  and 


no  Vacuum  Tube  Amplifiers 


THE  RADIO 


Figure  4 

STANDARD  CIRCUIT  FOR  RESISTANCE- 
CAPACITANCE  COUPLED  TRIODE  AM- 
PLIFIER STAGE 


pentodes  are  used;  mode  amplifier  stages 
will  be  discussed  first. 

R-C  Coupled  Figure  4 illustrates  the  stand- 
Triode  Stages  ard  circuit  for  a resistance- 
capacitance  coupled  amplifier 
stage  utilizing  a triode  tube  with  cathode  hias. 
In  conventional  audio-frequency  amplifier  de- 
sign such  stages  ate  used  at  medium  voltage 


levels  (from  0.01  to  5 volts  peak  on  the  grid 
of  the  tube)  and  use  medium-p  triodes  such 
as  the  6]  5 or  high-p  triodes  such  as  the  6SF5 
or  6SL7-GT.  Normal  voltage  gain  for  a single 
stage  of  this  type  is  from  10  to  70,  depending 
upon  the  tube  chosen  and  its  operating  con- 
ditions. Triode  tubes  are  normally  used  in  the 
last  voltage  amplifier  stage  of  an  R-C  ampli- 
fier since  their  harmonic  distortion  with  large 
output  voltage  (25  to  75  volts)  is  less  than 
with  a pentode  tube. 

Voltage  Gain  The  voltage  gain  pet  stage  of 
per  Stage  a resistance-capacitance  cou- 
pled triode  amplifier  can  be  cal- 
culated with  the  aid  of  the  equivalent  circuits 
and  expressions  for  the  mid-frequency,  high- 
frequency,  and  low-frequency  range  given  in 
figure  5- 

A triode  R-C  coupled  amplifie;'  stage  is 
normally  operated  with  values  of  cathode  re- 
sistor and  plate  load  resistor  such  that  the 
actual  voltage  on  the  tube  is  approximately 
one-half  the  d-c  plate  supply  voltage.  To 


E = 


JU  Rl  R6 

Rp  (Rl+Rs)+Rl  Rs 


MID  FREQUENCY  RANGE 


; Cgk 

(DYNAMIC, 
NEXT  STAGE) 


HIGH  FREQUENCY  RANGE 


A HIGH  FREQ.  _ 1 

A MID  FREQ.  ^ ,+  (ReQ/Xs)2 


R EQ  = 


Rl 


Rl 

Rp 


Xs  = 


1 

ZrTF  (Cpk+Cgk  (dynamic) 


LOW  FREQUENCY  RANGE 


A LOW  FREQ.  1 

A MID  FREQ.  ^ , + (Xc/RP 


Xc 


1 

ZTT'fCc 


R 


=5  Rg  + 


Rl  Rp 
Rl+  Rp 


Figure  5 

Equivalent  circuits  and  gain  equations  for  a triode  R-C  coupled  amplifier  stage.  In  using  these 
equations,  he  sure  to  select  the  valuos  of  mu  and  Rp  which  are  proper  for  the  static  current  and 
voltages  with  which  the  tube  will  operate.  These  values  may  be  obtained  from  curves  published 

in  the  RCA  Tube  Handbook  RC-76. 


HANDBOOK 


R-C  Coupled  Amplifiers  111 


Figure  6 

STANDARD  CIRCUIT  FOR  RESISTANCE- 
CAPACITANCE  COUPLED  PENTODE  AM- 
PLIFIER STAGE 


assist  the  designer  of  such  stages,  data  on 
operating  conditions  for  commonly  used  tubes 
is  published  in  the  RCA  Tube  Handbook  RC- 16. 
It  is  assumed,  in  the  case  of  the  gain  equations 
of  figure  5,  that  the  cathode  by-pass  capacitor, 
C),,  has  a reactance  that  is  low  with  respect 
to  the  cathode  resistor  at  the  lowest  frequency 
to  be  passed  by  the  amplifier  stage. 

R-C  Coupled  Figure  6 illustrates  the  stand- 
Pentode  Stoges  ard  circuit  for  a resistance- 
capacitance  coupled  pentode 
amplifier  stage.  Cathode  bias  is  used  and  the 
screen  voltage  is  supplied  through  a dropping 
resistor  from  the  plate  voltage  supply.  In  con- 
ventional audio-frequency  amplifier  design 
such  stages  are  normally  used  at  low  voltage 
levels  (from  0.00001  to  0.1  volts  peak  on  the 
grid  of  the  tube)  and  use  modetate-Gm  pentodes 


such  as  the  6SJ7.  Normal  voltage  gain  for  a 
stage  of  this  type  is  from  60  to  250,  depend- 
ing upon  the  tube  chosen  and  its  operating 
conditions.  Pentode  tubes  are  ordinarily  used 
the  first  stage  of  an  R-C  amplifier  where  the 
high  gain  which  they  afford  is  of  greatest  ad- 
vantage and  where  only  a small  voltage  output 
is  required  from  the  stage. 

The  voltage  gain  per  stage  of  a resistance- 
capacitance  coupled  pentode  amplifier  can  be 
calculated  with  the  aid  of  the  equivalent  cir- 
cuits and  expressions  for  the  mid-frequency, 
high-frequency,  and  low-frequency  range  given 
in  figure  7. 

To  assist  the  designer  of  such  stages,  data 
on  operating  conditions  for  commonly  used 
types  of  tubes  is  published  in  the  RCA  Tube 
Handbook  RC-16.  It  is  assumed,  in  the  case  of 
the  gain  equations  of  figure  7,  that  the  cathode 
by-pass  capacitor,  Cj;,  has  a reactance  that  is 
low  with  respect  to  the  cathode  resistor  at  the 
lowest  frequency  to  be  passed  by  the  stage.  It 
is  additionally  assumed  that  the  reactance  of 
the  screen  by-pass  capacitor  Cd,  is  low  with 
respect  to  the  screen  dropping  resistor,  Rd,  at 
the  lowest  frequency  to  be  passed  by  the  am- 
plifier stage. 

Coscade  Voltage  When  voltage  amplifier  stages 
Amplifier  Stages  are  operated  in  such  a manner 
that  the  output  voltage  of  the 
first  is  fed  to  the  grid  of  the  second,  and  so 
forth,  such  stages  are  said  to  be  cascaded. 
The  total  voltage  gain  of  cascaded  amplifier 


Figure  7 

Equivalent  circuits  and 
gain  equations  for  a pen- 
tode R-C  coupled  amplifier 
stage.  In  using  these  equa- 
tions be  sure  to  select  the 
values  of  G,„  and  Rp  which 
are  proper  for  the  static 
currents  and  voltages  with 
which  the  tube  will  oper- 
ate. These  values  may  be 
obtained  from  curves  pub- 
lished in  the  RCA  Tube 
Handbook  RC-16. 


|Rp 


MID  FREQUENCY  RANGE 


n 

- 5 

:Rl  «Rc  ^ 

"1 

1 

CckCdynamic) 


HIGH  FREQUENCY  RANGE 


1=  -Gm  Ec  ^|Cc 


Sro 


LOW  FREQUENCY  RANGE 


A = Gm  ReQ 


Req 


1 + 


Rl 

Rg  Rp 


A HIGH  FREQ-  _ 1 

A MIO  FREQ.  ^ 1+  (ReQ/Xs)2 


Req  = 


^ ^ Rg  Rp 


Xs  = 


1 

ZTTf  CCpk+Cgk  (dynamic) 


A LOW  FREQ.  _ 1 

A MID  FREQ.  ^TTTxcTrTz 


XC  = 


1 

ZJTfCc 

Rl ^ 


R=  Rg  + 


Rl+  Rp 


112  Vacuum  Tube  Amplifiers 


THE  RADIO 


Figure  8 

The  variation  of  stage  gain  with  frequency 
in  an  r-c  coupled  pentode  amplifier  for  vari- 
ous values  of  plate  load  resistance 


Stages  is  obtained  by  taking  the  product  of  the 
voltage  gains  of  each  of  the  successive  stages. 

Sometimes  the  voltage  gain  of  an  amplifier 
stage  is  rated  in  decibels.  Voltage  gain  is 
converted  into  decibels  gain  through  the  use 
of  the  following  expression:  db  = 20  logjo  A, 
where  A is  the  voltage  gain  of  the  stage.  The 
total  gain  of  cascaded  voltage  amplifier  stages 
can  be  obtained  by  adding  the  number  of 
decibels  gain  in  each  of  the  cascaded  stages. 

R-C  Amplifier  A typical  frequency  response 
Response  curve  for  an  R-C  coupled  audio 

amplifier  is  shown  in  figure  8. 
It  is  seen  that  the  amplification  is  poor  for  the 
extreme  high  and  low  frequencies.  The  reduced 
gain  at  the  low  frequencies  is  caused  by  the 
loss  of  voltage  across  the  coupling  capacitor. 
In  some  cases,  a low  value  of  coupling  capaci- 
tor is  deliberately  chosen  to  reduce  the  re- 
sponse of  the  stage  to  hum,  or  to  attenuate 
the  lower  voice  frequencies  for  communication 
purposes.  For  high  fidelity  work  the  product  of 
the  grid  resistor  in  ohms  times  the  coupling 
capacitor  in  microfarads  should  equal  25,000. 
(ie.:  500,000  ohms  x 0.05  liid  = 25,000). 

The  amplification  of  high  frequencies  falls 
off  because  of  the  Miller  effect  of  the  sub- 
sequent stage,  and  the  shunting  effect  of  resi- 
dual circuit  capacities.  Both  of  these  effects 
may  be  minimized  by  the  use  of  a low  value  of 
plate  load  resistor. 


Grid  Leak  Bias  The  correct  operating  bias 
(or  High  Mu  Triodes  for  a high-mu  triode  such 
as  the  6SL7,  is  fairly  crit- 
ical, and  will  be  found  to  be  highly  variable 
from  tube  to  tube  because  of  minute  variations 
in  contact  potential  within  the  tube  itself.  A 
satisfactory  bias  method  is  to  use  grid  leak 
bias,  with  a grid  resistor  of  one  to  ten  meg- 


MIO-FREQUCNCV  GAIN  = &MVi  Rl 

HIGH-FRCQUENCV  CAIN  = GmVi  Z cOUPLi  NG  NET  WORK 

C = CoUT  Vi  -P  C iN^a  ■(- c PISTRIBUTED 
FOR  COMPROMISE  HIGH  FREQUENCY  EQUALIZATION. 

Xll=  0.5  Xc  at  fc 
Rl  = XC  AT  fc 

WHERE:  Fc  = CUTOFF  FREQUENCY  OF  AMPLIFIER 
Ll  = PEAKING  INDUCTOR 
FOR  COMPROMISE  LOW  FREQUENCY  EQUALIZATION 

Rb  - Rk  (Gm^'i  Rl) 

Rb  Cb * Rk Ck 

Ck  - 25  TO  50  JJFO.  IN  PARALLEL  WITH  .001  MIC* 

Cb*  capacitance  from  ABOVE  WITH  001  MICA  IN  PARALLEL 

Figure  9 

SIMPLE  COMPENSATED  VIDEO 
AMPLIFIER  CIRCUIT 

Resistor  in  con/uncfi'on  with  coil 

serves  to  flatten  the  high-frequency  response 
of  the  stage,  while  and  R^  serve  to  equal- 
ize the  low-frequency  response  of  this  sim- 
ple video  amplifier  stage. 


ohms  connected  directly  between  grid  and 
cathode  of  the  tube.  The  cathode  is  grounded. 
Grid  current  flows  at  all  times,  and  the  effec- 
tive input  resistance  is  about  one-half  the 
resistance  value  of  the  grid  leak.  This  circuit 
is  particularly  well  suited  as  a high  gain 
amplifier  following  low  output  devices,  such 
as  crystal  microphones,  or  dynamic  micro- 
phones. 

R-C  Amplifier  A resistance-capacity 

General  Choracteristics  coupled  amplifier  can 
be  designed  to  provide 
a good  frequency  response  for  almost  any 
desired  range.  For  instance,  such  an  amplifier 
can  be  built  to  provide  a fairly  uniform  ampli- 
fication for  frequencies  in  the  audio  range  of 
about  100  to  20,000  cycles.  Changes  in  the 
values  of  coupling  capacitors  and  load  re- 
sistors can  extend  this  frequency  range  to 
cover  the  very  wide  range  required  for  video 
service.  However,  extension  of  the  range  can 
only  be  obtained  at  the  cost  of  reduced  over- 
all amplification.  Thus  the  R-C  method  of 
coupling  allows  good  frequency  response  with 
minimum  distortion,  but  low  amplification. 
Phase  distortion  is  less  with  R-C  coupling 


HANDBOOK 


Video  Frequency  Amplifiers  113 


than  with  other  types,  except  direct  coupling. 
The  R'C  amplifier  may  exhibit  tendencies  to 
"motorboat”  or  oscillate  if  it  is  used  with  a 
high  impedance  plate  supply. 


6-6  Video-Frequency 

Amplifiers 

A video-frequency  amplifier  is  one  which 
has  been  designed  to  pass  frequencies  from 
the  lower  audio  range  (lower  limit  perhaps  50 
cycles)  to  the  middle  r-f  range  (upper  limit 
perhaps  4 to  6 megacycles).  Such  amplifiers, 
in  addition  to  passing  such  an  extremely  wide 
frequency  range,  must  be  capable  of  amplify- 
ing this  range  with  a minimum  of  amplitude, 
phase,  and  frequency  distortion.  Video  ampli- 
fiers are  commonly  used  in  television,  pulse 
communication,  and  radar  work. 

Tubes  used  in  video  amplifiers  must  have 
a high  ratio  of  G„,  to  capacitance  if  a usable 
gain  per  stage  is  to  be  obtained.  Commonly 
available  tubes  which  have  been  designed  for 
or  are  suitable  for  use  in  video  amplifiers  are; 
6AU6,  6AG5,  6AK5,  6CB6,  6AC7,  6AG7,  and 
6K6-GT.  Since,  at  the  upper  frequency  limits 
of  a video  amplifier  the  input  and  output 
shunting  capacitances  of  the  amplifier  tubes 
have  rather  low  values  of  reactance,  low 
values  of  coupling  resistance  along  with 
peaking  coils  ot  other  special  interstage  cou- 
pling impedances  are  usually  used  to  flatten 
out  the  gain/ftequency  and  hence  the  phase/ 
frequency  characteristic  of  the  amplifier. 
Recommended  operating  conditions  along  with 
expressions  for  calculation  of  gain  and  circuit 
values  are  given  in  figure  9.  Only  a simple 
two-terminal  interstage  coupling  network  is 
shown  in  this  figure. 

The  performance  and  gain-per- stage  of  a 
video  amplifier  can  be  improved  by  the  use 
of  increasingly  complex  two-terminal  inter- 
stage coupling  netwotks  or  through  the  use 
of  four-terminal  coupling  networks  or  filters 
between  successive  stages.  The  reader  is  re- 
ferred to  Terman’s  "Radio  Engineer's  Hand- 
book” for  design  data  on  such  interstage 
coupling  networks. 


6-7  Other  Interstage 

Coupling  Methods 

Figure  10  illustrates,  in  addition  to  resist- 
ance-capacitance interstage  coupling,  seven 
additional  methods  in  which  coupling  between 
two  successive  stages  of  an  audio-frequency 
amplifier  may  be  accomplished.  Although 


resistance-capacitance  coupling  is  most  com- 
monly used,  there  are  certain  circuit  condi- 
tions wherein  coupling  methods  other  than 
resistance  capacitance  are  more  effective. 

Transformer  Transformer  coupling,  as  illus- 
Coupling  trated  in  figure  lOB,  is  seldom 
used  at  the  present  time  between 
two  successive  single-ended  stages  of  an 
audio  amplifier.  There  are  several  reasons  why 
resistance  coupling  is  favored  over  transformer 
coupling  between  two  successive  single-ended 
stages.  These  are:  (1)  a transformer  having 
frequency  characteristics  comparable  with  a 
properly  designed  R-C  stage  is  very  expensive; 
(2)  transformers,  unless  they  are  very  well 
shielded,  will  pick  up  inductive  hum  from 
nearby  power  and  filament  transformers;  (3) 
the  phase  characteristics  of  step-up  interstage 
transformers  are  poor,  making  very  difficult 
the  inclusion  of  a transformer  of  this  type 
within  a feedback  loop;  and  (4)  transformers 
are  heavy. 

However,  there  is  one  circuit  application 
where  a step-up  interstage  transformer  is  of 
considerable  assistance  to  the  designer;  this 
is  the  case  where  it  is  desired  to  obtain  a 
large  amount  of  voltage  to  excite  the  grid  of  a 
cathode  follower  or  of  a high-power  Class  A 
amplifier  from  a tube  operating  at  a moderate 
plate  voltage.  Under  these  conditions  it  is  pos- 
sible to  obtain  apeak  voltage  on  the  secondary 
of  the  transformer  of  a value  somewhat  greater 
than  the  d-c  plate  supply  voltage  of  the  tube 
supplying  the  primary  of  the  transformer. 

Push-Pull  Transformer  Push-pull  transformer 

Interstage  Coupling  coupling  between  two 

stages  is  illustrated  in 
figure  IOC.  This  interstage  coupling  arrange- 
ment is  fairly  commonly  used.  The  system  is 
particularly  effective  when  it  is  desired,  as  in 
the  system  just  described,  to  obtain  a fairly 
high  voltage  to  excite  the  grids  of  a high- 
power  audio  stage.  The  arrangement  is  also 
very  good  when  it  is  desired  to  apply  feed- 
back to  the  grids  of  the  push-pull  stage  by 
applying  the  feedback  voltage  to  the  low- 
potential  sides  of  the  two  push-pull  second- 
aries. 

Impedance  Impedance  coupling  between  two 
Coupling  stages  is  shown  in  figure  lOD. 

This  circuit  arrangement  is  seldom 
used,  but  it  offers  one  strong  advantage  over 
R-C  interstage  coupling.  This  advantage  is 
the  fact  that,  since  the  operating  voltage  on 
the  tube  with  the  impedance  in  the  plate  cir- 
cuit is  the  plate  supply  voltage,  it  is  possible 
to  obtain  approximately  twice  the  peak  volt- 
age output  that  it  is  possible  to  obtain  with 
R-C  coupling.  This  is  because,  as  has  been 


114  Vacuum  Tube  Amplifiers 


THE  RADIO 


@ RESISTANCE-CAPACITANCE  COUPLING  (S)  TRANSFORMER  COUPLING 


@ PUSH-PULL  TRANSFORMER  COUPLING  ® IMPEDANCE  COUPLING 


© CATHODE  COUPLING  ® DIRECT  COUPLING 


Figure  10 

INTERSTAGE  COUPLING  METHODS  FOR  AUDIO  FREQUENCY  VOLTAGE  AMPLIFIERS 


mentioned  before,  the  d-c  plate  voltage  on  an 
R-C  stage  is  approximately  one-half  the  plate 
supply  voltage. 

Impedance-Transformer  These  two  circuit  ar- 
and  Resistance-Trans-  rangements,  illustrated 
former  Coupling  in  figures  lOE  and  lOF, 

are  employed  when  it  is 
desired  to  use  transformer  coupling  for  the 
reasons  cited  above,  but  where  it  is  desired 
that  the  d-c  plate  current  of  the  amplifier 


stage  be  isolated  from  the  primary  of  the  cou- 
pling transformer.  With  most  types  of  high- 
permeability  wide-tesponse  transformers  it  is 
necessary  that  there  be  no  direct-current  flow 
through  the  windings  of  the  transformer.  The 
impedance-transformer  arrangement  of  figure 
lOE  will  give  a higher  voltage  output  from 
the  stage  but  is  not  often  used  since  the  plate 
coupling  impedance  (choke)  must  have  very 
high  inductance  and  very  low  distributed  ca- 
pacitance in  order  not  to  restrict  the  range  of 


HANDBOOK 


Phase  Inverters  115 


Gm'=-Gm  ^ - G=  RkGu  O + TT) 

2G+1 

Rk  • CATHODE  RESISTOR 

Rp'  = Rp  — Gm  = Gm  of  each  tube 

GTI 

JU  = JU  OF  EACH  TUBE 
W = -U  Rp  = Rp  OF  EACH  TUBE 

EQUIVALENT  FACTORS  INDICATED  ABOVE  BY  (O  ARE 
THOSE  OBTAINED  BY  USING  AN  AMPLIFIER  WITH  A PAIR 
OF  SIMILAR  TUBE  TYPES  IN  CIRCUIT  SHOWN  ABOVE. 

Figure  1 1 

Equivatenf  factors  for  a pair  of  similar  fri- 
odes  operating  as  a cathode^coupled  audios 
frequency  voltage  amplifier. 


the  transformer  which  it  and  its  associated 
tube  feed.  The  resistance-transformer  arrange- 
ment of  figure  10  F is  ordinarily  quite  satis- 
factory where  it  is  desired  to  feed  a trans- 
former from  a voltage  amplifier  stage  with  no 
d.c.in  the  transformer  primary. 

Cathode  The  cathode  coupling  arrangement 
Coupling  of  figure  lOG  has  been  widely  used 
only  comparatively  recently.  One 
outstanding  characteristic  of  such  a circuit  is 
that  there  is  no  phase  reversal  between  the 
grid  and  the  plate  circuit.  All  other  common 
types  of  interstage  coupling  are  accompanied 
by  a 180°  phase  reversal  between  the  grid 
circuit  and  Ae  plate  circuit  of  the  tube. 

Figure  11  gives  the  expressions  for  deter- 
mining the  appropriate  factors  for  an  equiva- 
lent triode  obtained  through  the  use  of  a pair 
of  similar  triodes  connected  in  the  cathode- 
coupled  circuit  shown.  With  these  equivalent 
triode  factors  it  is  possible  to  use  the  ex- 
pressions shown  in  figure  5 to  determine  the 
gain  of  the  stage  at  different  frequencies.  The 
input  capacitance  of  such  a stage  is  less  than 
that  of  one  of  the  triodes,  the  effective  grid- 
to-plate  capacitance  is  very  much  less  (it  is 
so  much  less  that  such  a stage  may  be  used 
as  an  r-f  amplifier  without  neutralization),  and 
the  output  capacitance  is  approximately  equal 
to  the  grid-to-plate  capacitance  of  one  of  the 
triode  sections.  This  circuit  is  particularly 
effective  with  tubes  such  as  the  6J6,  6N7,  and 
6SN7-GT  which  have  two  similar  triodes  in 
one  envelope.  An  appropriate  value  of  cathode 
resistor  to  use  for  such  a stage  is  the  value 
which  would  be  used  for  die  cathode  resistor 
of  a conventional  amplifier  using  one  of  the 


same  type  tubes  with  the  values  of  plate  volt- 
age and  load  resistance  to  be  used  for  the 
cathode-coupled  stage. 

Inspection  of  the  equations  in  figure  11 
shows  that  as  the  cathode  resistor  is  made 
smaller,  to  approach  zero,  the  G^  approaches 
zero,  the  plate  resistance  approaches  the  Rp 
of  one  tube,  and  the  mu  approaches  zero.  As 
the  cathode  resistor  is  made  very  large  the  Gn, 
approaches  one  half  that  of  a single  tube  of 
the  same  type,  the  plate  resistance  approaches 
twice  that  of  one  tube,  and  the  mu  approaches 
the  same  value  as  one  tube.  But  since  the  Gn 
of  each  tube  decreases  as  the  cathode  resistor 
is  made  larger  (since  the  plate  current  will 
decrease  on  each  tube)  the  optimum  value  of 
cathode  resistor  will  be  found  to  be  in  the 
vicinity  of  the  value  mentioned  in  the  previous 
paragraph. 

Direct  Coupling  Direct  coupling  between  suc- 
cessive amplifier  stages  (plate 
of  first  stage  connected  directly  to  the  grid  of 
the  succeeding  stage)  is  complicated  by  the 
fact  that  the  grid  of  an  amplifier  stage  must 
be  operated  at  an  average  negative  poten- 
tial with  respect  to  the  cathode  of  that  stage. 
However,  if  the  cathode  of  the  second  ampli- 
fier stage  can  be  operated  at  a potential  more 
positive  than  the  plate  of  the  preceding  stage 
by  the  amount  of  the  grid  bias  on  the  second 
amplifier  stage,  this  direct  connection  between 
the  plate  of  one  stage  and  the  grid  of  the  suc- 
ceeding stage  can  be  used.  Figure  lOH  il- 
lustrates an  application  of  this  principle  in 
the  coupling  of  a pentode  amplifier  stage  to 
the  grid  of  a "hot-cathode”  phase  inverter.  In 
this  arrangement  the  values  of  cathode,  screen, 
and  plate  resistor  in  the  pentode  stage  are 
chosen  such  that  the  plate  of  the  pentode  is  at 
approximately  0.  3 times  the  plate  supply  po- 
tential. The  succeeding  phase-inverter  stage 
then  operates  with  conventional  values  of 
cathode  and  plate  resistor  (same  value  of  re- 
sistance) in  its  normal  manner.  This  type  of 
phase  inverter  is  described  in  more  detail  in 
the  section  to  follow. 


6-8  Phase  Inverters 

It  is  necessary  in  order  to  excite  the  grids 
of  a push-pull  stage  that  voltages  equal  in 
amplitude  and  opposite  in  polarity  be  applied 
to  the  two  grids.  These  voltages  may  be  ob- 
tained through  the  use  of  a push-pull  input 
transformer  such  as  is  shown  in  figure  IOC. 
It  is  possible  also,  without  the  attendant  bulk 
and  expense  of  a push-pull  input  transformer, 
to  obtain  voltages  of  the  proper  polarity  and 


116  Vacuum  Tube  Amplifiers 


THE  RADIO 


phase  through  the  use  of  a so-called  phase- 
inverter  stage.  There  ate  a large  number  of 
phase  inversion  circuits  which  have  been  de- 
veloped and  applied  hut  the  three  shown  in 
figure  12  have  been  found  over  a period  of 
time  to  be  the  most  satisfactory  from  the  point 
of  view  of  the  number  of  components  requited 
and  from  the  standpoint  of  the  accuracy  with 
which  the  two  out-of-phase  voltages  are  held 
to  the  same  amplitude  with  changes  in  supply 
voltage  and  changes  in  tubes. 

All  of  these  vacuum  tube  phase  inverters 
ate  based  upon  the  fact  that  a 180°  phase 
shift  occurs  within  a vacuum  tube  between  the 
grid  input  voltage  and  the  plate  output  voltage. 
In  certain  circuits,  the  fact  that  the  grid  input 
voltage  and  the  voltage  appearing  across  the 
cathode  bias  resistor  are  in  phase  is  used  for 
phase  inversion  purposes. 

"Hot-Cathode"  Figure  12A  illustrates  the  hot- 
Phose  Inverter  cathode  type  of  phase  in- 
verter. This  type  of  phase  in- 
verter is  the  simplest  of  the  three  types  since 
it  requires  only  one  tube  and  a minimum  of 
circuit  components.  It  is  particularly  simple 
when  directly  coupled  from  the  plate  of  a 
pentode  amplifier  stage  as  shown  in  figure 
lOH.  The  circuit  does,  however,  possess  the 
following  two  disadvantages:  (1)  the  cathode 
of  the  tube  must  run  at  a potential  of  approxi- 
mately 0.3  times  the  plate  supply  voltage 
above  the  heater  when  a grounded  common 
heater  winding  is  used  for  this  tube  as  well 
as  the  other  heater-cathode  tubes  in  a receiver 
or  amplifier:  (2)  the  circuit  actually  has  a 
loss  in  voltage  from  its  input  to  either  of  the 
output  grids  — about  0.9  times  the  input  volt- 
age will  be  applied  to  each  of  these  grids. 
This  does  represent  a voltage  gain  of  about 
1.8  in  total  voltage  output  with  respect  to  in- 
put (grid-to-grid  output  voltage)  but  it  is  still 
small  with  respect  to  the  other  two  phase 
inverter  circuits  shown. 

Recommended  component  values  for  use 
with  a 6J5  tube  in  this  circuit  are  shown  in 
figure  12A.  If  it  is  desired  to  use  another  tube 
in  this  circuit,  appropriate  values  for  the  oper- 
ation of  that  tube  as  a conventional  amplifier 
can  be  obtained  from  manufacturer’s  tube  data. 
The  value  of  obtained  should  be  divided  by 
two,  and  this  new  value  of  resistance  placed 
in  the  circuit  as  Rl-  The  value  of  Rj;  from 
tube  manual  tables  should  then  be  used  as 
R);!  in  this  circuit,  and  then  the  total  of  R|(i 
and  Ri;2  should  be  equal  to  R^. 

"Floating  Porophose”  An  alternate  type  of 
Phase  Inverter  phase  inverter  some- 

times called  the  "float- 
ing paraphase”  is  illustrated  in  figure  12B. 
This  circuit  is  quite  often  used  with  a 6N7 


@ "HOT  CATHODE"  PHASE  INVERTER 


© CATHODE  COUPLED  PHASE  INVERTER 


Figure  12 

THREE  POPULAR  PHASE-INVERTER  CIR- 
CUITS WITH  RECOMMENDED  VALUES  FOR 
CIRCUIT  COMPONENTS 


tube,  and  appropriate  values  for  the  6N7  tube 
in  this  application  are  shown.  The  circuit 
shown  with  the  values  given  will  give  a volt- 
age gain  of  approximately  21  from  the  input 
grid  to  each  of  the  grids  of  the  succeeding 
stage.  It  is  capable  of  approximately  70  volts 
peak  output  to  each  grid. 

The  circuit  inherently  has  a small  unbalance 
in  output  voltage.  This  unbalance  can  be  elim- 
inated, if  it  is  requited  for  some  special  appli- 
cation, by  making  the  resistor  Rgi  a few  per 
cent  lower  in  resistance  value  than  Rg3. 

Cathode-Coupled  The  circuit  shown  in  figure 
Phase  Inverter  12C  gives  approximately  one- 
half  the  voltage  gain  from  the 
input  grid  to  either  of  the  grids  of  the  suc- 
ceeding stage  that  would  be  obtained  from  a 
single  tube  of  the  same  type  operating  as  a 
conventional  R-C  amplifier  stage.  Thus,  with 
a 6SN7-GT  tube  as  shown  (two  6J5’s  in  one 


HANDBOOK 


Vacuum  Tube  Voltmeter  117 


.01  RS  R6 


INVERTER 


envelope)  the  voltage  gain  from  the  input 
grid  to  either  of  the  output  grids  will  be  ap- 
proximately 7 — the  gain  is,  of  course,  14  from 
the  input  to  both  output  grids.  The  phase 
characteristics  are  such  that  the  circuit  is 
commonly  used  in  deriving  push-pull  deflec- 
tion voltage  for  a cathode-ray  tube  from  a 
signal  ended  input  signal. 

The  first  half  of  the  6SN7  is  used  as  an 
amplifier  to  increase  the  amplitude  of  the  ap- 
plied signal  to  the  desired  level.  The  second 
half  of  the  6SN7  is  used  as  an  inverter  and 
amplifier  to  produce  a signal  of  the  same 
amplitude  but  of  opposite  polarity.  Since  the 
common  cathode  resistor,  R|;,  is  not  by-passed 
the  voltage  across  it  is  the  algebraic  sum  of 
the  two  plate  currents  and  has  the  same  shape 
and  polarity  as  the  voltage  applied  to  the  in- 
put grid  of  the  first  half  of  the  6SN7.  When  a 
signal,  e,  is  applied  to  the  input  circuit,  the 
effective  grid-cathode  voltage  of  the  first 
section  is  Ae/2,  when  A is  the  gain  of  the 
first  section.  Since  the  grid  of  the  second 
section  of  the  6SN7  is  grounded,  the  effect  of 
the  signal  voltage  across  R|;  (equal  to  e/2  if 
R);  is  the  proper  value)  is  the  same  as  though 
a signal  of  the  same  amplitude  but  of  opposite 
polarity  were  applied  to  the  grid.  The  output 
of  the  second  section  is  equal  to  - Ae/2  if  the 
plate  load  resistors  are  the  same  for  both  tube 
sections. 


Voltage  Divider  A commonly  used  phase  in- 
Phose  Inverter  vetter  is  shown  in  figure  13. 

The  input  section  (V,)  is  con- 
nected as  a conventional  amplifier.  The  out- 
put voltage  from  V,  is  impressed  on  the  volt- 
age divider  R5-R5.  The  values  of  Rj  and  R5 
are  in  such  a ratio  that  the  voltage  impressed 
upon  the  grid  of  Yj  is  1/A  times  the  output 
voltage  of  Vi,  where  A is  the  amplification 
factor  of  V,.  The  output  of  Vj  is  then  of  the 


Figure  14 

DIRECT  COUPLED 
D-C  AMPLIFIER 


same  amplitude  as  the  output  of  V,,  but  of 
opposite  phase. 


6-9  D-C  Amplifiers 

Direct  current  amplifiers  are  special  types 
used  where  amplification  of  very  slow  varia- 
tions in  voltage,  or  of  d-c  voltages  is  desired. 
A simple  d-c  amplifier  consists  of  a single 
tube  with  a grid  resistor  across  the  input 
terminals,  and  the  load  in  the  plate  circuit. 


Bosic  D-C  A simple  d-c  amplifier 

Amplifier  Circuit  circuit  is  shown  in 

figure  14,  wherein  the 
the  grid  of  one  tube  is  connected  directly  to 
the  plate  of  the  preceding  tube  in  such  a 
manner  that  voltage  changes  on  the  grid  of 
the  first  tube  will  be  amplified  by  the  system. 
The  voltage  drop  across  the  plate  coupling 
resistor  is  impressed  directly  upon  the  grid 
of  the  second  tube,  which  is  provided  with 
enough  negative  grid  bias  to  balance  out  the 
excessive  voltage  drop  across  the  coupling 
resistor.  The  grid  of  the  second  tube  is  thus 
maintained  in  a slightly  negative  position. 

The  d-c  amplifier  will  provide  good  low  fre- 
quency response,  with  negligible  phase  dis- 
tortion. High  frequency  response  is  limited 
by  the  shunting  effect  of  the  tube  capacitances, 
as  in  the  normal  resistance  coupled  amplifier. 

A common  fault  with  d-c  amplifiers  of  all 
types  is  static  instability.  Small  changes  in 
the  filament,  plate,  or  grid  voltages  cannot 
be  distinguished  from  the  exciting  voltage. 
Regulated  power  supplies  and  special  balanc- 
ing circuits  have  been  devised  to  reduce  the 
effects  of  supply  variations  on  these  ampli- 
fiers. A successful  system  is  to  apply  the 
place  potential  in  phase  to  two  tubes,  and  to 
apply  the  exciting  signal  to  a push-pull  grid 


118  Vacuum  Tube  Amplifiers 


THE  RADIO 


PUSH-PULL  D-C  AMPLIFIER 
WITH  EITHER  SINGLE-ENDED 
OR  PUSH-PULL  INPUT 


circuit  configuration.  If  the  two  tubes  ate 
identical,  any  change  in  electrode  voltage  is 
balanced  out.  The  use  of  negative  feedback 
can  also  greatly  reduce  drift  problems. 

The  “Loftin-White”  Two  d-c  amplifier 

Circuit  stages  may  be  arranged, 

so  that  their  plate 
supplies  are  effectively  in  series,  as  illus- 
trated in  figure  15-  This  is  known  as  a Loftin- 
V/hite  amplifier.  All  plate  and  grid  voltages 
may  be  obtained  from  one  master  power  supply 
instead  of  separate  grid  and  plate  supplies. 
A push-pull  version  of  this  amplifier  (figure  16) 
can  be  used  to  balance  out  the  effects  of  slow 
variations  in  the  supply  voltage. 


6-10  Single-ended  Triode 
Amplifiers 

Figure  17  illustrates  five  circuits  for  the 
operation  of  Class  A triode  amplifier  stages. 
Since  the  cathode  current  of  a triode  Class  Aj 
(no  grid  current)  amplifier  stage  is  constant 
with  and  without  excitation,  it  is  common 
practice  to  operate  the  tube  with  cathode 
bias.  Recommended  operating  conditions  in 
regard  to  plate  voltage,  grid  bias,  and  load 
impedance  for  conventional  triode  amplifier 
stages  are  given  in  the  RCA  Tube  Manual, 
RC-16. 


Extended  Closs  A It  is  possible,  under  certain 
Operation  conditions  to  operate  single- 

ended  triode  amplifier  stages 
(and  pentode  and  tetrode  stages  as  well)  with 
grid  excitation  of  sufficient  amplitude  that 
grid  current  is  taken  by  the  tube  on  peaks. 
This  type  of  operation  is  called  Class  A2  and 


is  characterized  by  increased  plate-circuit 
efficiency  over  straight  Class  A amplification 
without  grid  current.  The  normal  Class  Aj 
amplifier  power  stage  will  operate  with  a plate 
circuit  efficiency  of  from  20  percent  to  perhaps 
35  per  cent.  Through  the  use  of  Class  A2 
operation  it  is  possible  to  increase  this  plate 
circuit  efficiency  to  approximately  38  to  45 
per  cent.  However,  such  operation  requires 
careful  choice  of  the  value  of  plate  load  im- 
pedance, a grid  bias  supply  with  good  regula- 
tion (since  the  tube  draws  grid  current  on 
peaks  although  the  plate  current  does  not 
change  with  signal),  and  a driver  tube  with 
moderate  power  capability  to  excite  the  grid 
of  the  Class  Aj  tube. 

Figures  17D  and  17E  illustrate  two  methods 
of  connection  for  such  stages.  Tubes  such  as 
the  845,  849,  and  304TL  ate  suitable  for  such 
a stage.  In  each  case  the  grid  bias  is  approxi- 
mately the  same  as  would  be  used  for  a Class 
Aj  amplifier  using  the  same  tube,  and  as 
mentioned  before,  fixed  bias  must  be  used 
along  with  an  audio  driver  of  good  regulation  — 
preferably  a triode  stage  with  a 1:1  or  step- 
down  driver  transformer.  In  each  case  it  will 
be  found  that  the  correct  value  of  plate  load 
impedance  will  be  increased  about  40  per  cent 
over  the  value  recommended  by  the  tube  manu- 
facturer for  Class  Aj  operation  of  the  tube. 

Operation  Character-  A Class  A power  amplifier 
istics  of  o Triode  operates  in  such  a way  as 
Power  Amplifier  to  amplify  as  faithfully  as 

possible  the  waveform  ap- 
plied to  the  grid  of  the  tube.  Large  power  out- 
put is  of  more  importance  than  high  voltage 
amplification,  consequently  gain  character- 
istics may  be  sacrificed  in  power  tube  design 
to  obtain  mote  important  power  handling  capa- 
bilities. Class  A power  tubes,  such  as  the  45, 
2A3  and  6AS7  are  characterized  by  a low 
amplification  factor,  high  plate  dissipation 
and  relatively  high  filament  emission. 

The  operating  characteristics  of  a Class  A 


HANDBOOK 


Triode  Amplifier  Characteristics  119 


® TRANSFORMER  COUPLING 


© IMPEDANCE-TRANSFORMER  COUPLING 


® TRANSFORMER  COUPLING  FOR  Aj  OPERATION 


© CLASS  Aj  MODULATOR  WITH  AUTO-TRANS- 
FORMER COUPLING 


Figure  17 

Output  coupling  arrangements  for  s/ng/e>en(/e</ 
Class  A triode  audlO’frequency  power  amplifiers. 


triode  amplifier  employing  an  output  trans- 
former-coupled load  may  be  calculated  from 
the  plate  family  of  curves  for  the  particular 
tube  in  question  by  employing  the  following 
steps: 

1-  The  load  tesistance  should  be  approxi- 
mately twice  the  plate  tesistance  of  the 
tube  for  maximum  undistorted  power  out- 
put. Remember  this  fact  for  a quick  check 
on  calculations. 

2-  Calculate  the  zero-signal  bias  voltage 

(Eg). 


-(0.  68  X Ebb) 

Eg  

Where  Ebb  is  the  actual  plate  voltage  of 
the  Class  A stage,  and  p is  the  amplifi- 
cation factor  of  the  tube. 

3-  Locate  the  Eg  bias  point  on  the  Ip  vs, 
Ep  graph  where  the  Eg  bias  line  crosses 
the  plate  voltage  line,  as  shown  in  figure 
18.  Call  this  point  P. 

4-  Locate  on  the  plate  family  of  curves  the 
value  of  zero-signal  plate  current.  Ip, 
corresponding  to  the  operating  point,  P. 

5-  Locate  2 x Ip  (twice  the  value  of  Ip)  on 
the  plate  current  axis  (Y-axis).  This 
point  corresponds  to  the  value  of  maxi- 
mum signal  plate  current,  i„)ax. 

6-  Locate  point  x on  the  d-c  bias  curve  at 
zero  volts  (Eg  = 0),  corresponding  to  the 
value  of 

7-  Draw  a straight  Iine(x  - y)  through  points 
X and  P.  This  line  is  the  load  resistance 
line.  Its  slope  corresponds  to  the  value 
of  the  load  resistance. 

8-  Load  Resistance,  (in  ohms) 


where  e is  in  volts,  i is  in  amperes,  and 
Rl  is  in  ohms. 

9-  Check:  Multiply  the  zero-signal  plate 
current.  Ip,  by  the  operating  plate  volt- 
age, Ep.  If  the  plate  dissipation  rating 
of  the  tube  is  exceeded,  it  is  necessary 
to  increase  the  bias  (Eg)  on  the  tube  so 
that  the  plate  dissipation  falls  within 
the  maximum  rating  of  the  tube.  If  this 
step  is  taken,  operations  2 through  8 
must  be  repeated  with  the  new  value  of 
Eg. 

10-  For  maximum  power  output,  the  peak  a-c 
grid  voltage  on  the  tube  should  swing  to 
2Eg  on  the  negative  cycle,  and  to  zero- 
bias  on  the  positive  cycle.  At  the  peak 
of  the  negative  swing,  the  plate  volt- 
age teaches  e^pg,;  and  the  plate  current 
drops  to  iipip.  On  the  positive  swing  of 
the  grid  signal,  the  plate  voltage  drops 
^min  the  plate  current  reaches 

tmax-  The  power  output  of  the  tube  is: 
Power  Output  (watts) 


8 


where  i is  in  amperes  and  e is  in  volts. 

11-  The  second  harmonic  distortion  generated 
in  n single-ended  Class  A triode  ampli- 
fier, expressed  as  a percentage  of  the 
fundamental  output  signal  is: 


THE  RADIO 


120  Vacuum  Tube  Amplifiers 


AVERAGE  PLATE  CHARACTERISTICS  - 2A3 

Rp=  800  OHMS 
PLATE  DISSIPATION  = 15  WATTS 


LOAD  RESISTANCE 


R|^  = ^MAX.—  Emin. 
I MAX.—  LmiN, 


OHMS 


POWER  OUTPUT 


Po 


(Imax,  ~Imin.)  ^Emax~EminJ 
8 


WATTS 


1 

o 

1 

oZs 

Rl« 

o 

] 

o 

V 

Figure  19 

Normal  singlo-onded  pentode  or  iieom  tetrode 
ou(//o-frequenc/  power  oufpuf  stoge. 


plifier  stage.  Tubes  of  this  type  have  largely 
replaced  triodes  in  the  output  stage  of  re- 
ceivers and  amplifiers  due  to  the  higher  plate 
efficiency  (30%— 40%)  with  which  they  operate. 
Tetrode  and  pentode  tubes  do,  however,  intro- 
duce a considerably  greater  amount  of  harmonic 
distortion  in  their  output  circuit,  particularly 
odd  harmonics.  In  addition,  their  plate  circuit 
impedance  (which  acts  in  an  amplifier  to  damp 
loudspeaker  overshoot  and  ringing,  and  acts 
in  a driver  stage  to  provide  good  regulation)  is 
many  times  higher  than  that  of  an  equivalent 
triodc.  The  application  of  negative  feedback 
acts  both  to  reduce  distortion  and  to  reduce 
the  effective  plate  circuit  impedance  of  these 
tubes. 


SECOND  harmonic  DISTORTION 
(Imax.+  Imin.)  “Ip 


Da 


2 

Imax.  — Imin, 


X 100  PERCENT 


Figure  18 

Formulas  for  determining  the  operating  con- 
ditions for  a Class  A triode  single-ended  audio- 
frequency power  output  stage.  A typical  toad 
line  has  been  drawn  on  the  average  plate  char- 
acteristics of  a type  2A3  tube  to  illustrate 
the  procedure. 


Operating  Character-  The  Operating  character- 
istics of  a Pentode  istics  of  pentode  power 
Power  Amplifier  amplifiers  may  be  obtained 

from  the  plate  family  of 
curves,  much  as  in  the  manner  applied  to 
triode  tubes.  A typical  family  of  pentode  plate 
curves  is  shown  in  figure  20.  It  can  be  seen 
from  these  curves  that  the  plate  current  of  the 
tube  is  relatively  independent  of  the  applied 
plate  voltage,  but  is  sensitive  to  screen  volt- 
age.  In  general,  the  correct  pentode  load  re- 
sistance is  about 


% 2d  harmonic  - 

“ Imin) 


; ^ (X  100) 

^max  *“  ^min 

Figure  18  illustrates  the  above  steps  as 
applied  to  a single  Class  A 2A3  ampli- 
fier stage. 

6-11  Single-ended  Pentode 
Amplifiers 

Figure  19  illustrates  the  conventional  cir- 
cuit for  a single-ended  tetrode  or  pentode  am- 


0.  9 Ep 

J 

Ip 

and  the  power  output  is  somewhat  less  than 
Ep  X Ip 
2 ■ 

These  formulae  may  be  used  for  a quick  check 
on  more  precise  calculations.  To  obtain  the 
operating  parameters  for  Class  A pentode  am- 
plifiers, the  following  steps  are  taken: 

1-  The  imax  point  is  chosen  so  as  to  fall  on 
the  zero-bias  curve,  just  above  the 
**knee’*  of  the  curve  (point  A,  figure  20) . 

2-  A preliminary  operating  point,  P,  is  deter- 
mined by  the  intersection  of  the  plate 
voltage  line,  Ep,  and  the  line  of  iniax/2. 


PLATE  MILLIAMPERES 


HANDBOOK 


Push-Pull  Amplifiers 


121 


Figure  20 

GRAPHIC  DETERMINATION  OF  OPERAT- 
ING CHARACTERISTICS  OF  A PENTODE 
POWER  AMPLIFIER 

V is  the  negative  control  grid  voltage  of 
the  operating  point  P 


6-  The  power  output  is: 

Power  Output  (watts) 

^ (Imax  ~ Itnin)  + 1.41  (I^  “ ly)^  ^ 

32 

Where  1*  is  the  plate  current  at  the  point 
on  the  load  line  where  the  grid  voltage, 
eg,  is  equal  to:  Eg  - 0.  7 Eg;  and  where 
ly  is  the  plate  current  at  the  point  where 
eg  is  equal  to:  Eg  + 0.  7 Eg. 

7-  The  percentage  harmonic  distortion  is: 

% 2d  harmonic  distortion 

^max  “ *min  ~ 2 In 

= . ^ X 100 

^max  ~ Imin  + 1.41  (Ig  ~ ly) 

Where  Ip  is  the  static  plate  current  of 
of  the  tube. 

% 3d  harmonic  distortion 

Imax  “ Imin  ~ 1.41  (Ig  — ly) 

= - ^ L X 100 

^max  “ Imin  + 1.41  (Ig  — ly) 


The  grid  voltage  curve  that  this  point 
falls  upon  should  be  one  that  is  about 
the  value  of  Eg  requited  to  cut  the  plate 
current  to  a very  low  value  (Point  B). 
Point  B represents  i„i„  on  the  plate  cur- 
rent axis  (y-axis).  The  line  imax/2  should 
be  located  half-way  between  i,pgg  and 
^min* 

3"  A trial  load  line  is  constructed  about 
point  P and  point  A in  such  a way  that 
the  lengths  A-P  and  P-B  are  approxi- 
mately equal, 

4-  \?1ien  the  most  satisfactory  load  line  has 
been  determined,  the  load  resistance  may 
calculated: 


5-  The  operating  bias  (Eg)  is  the  bias  at 
point  P. 


6-12  Push-Pull  Audio 

Amplifiers 

A number  of  advantages  are  obtained  through 
the  use  of  the  push-pull  connection  of  two  or 
four  tubes  in  an  audio-frequency  power  am- 
plifier. Two  conventional  circuits  for  the  use 
of  triode  and  tetrode  tubes  in  the  push-pull 
connection  are  shown  in  figure  21.  The  two 
main  advantages  of  the  push-pull  circuit  ar- 
rangement are:  (1)  the  magnetizing  effect 
of  the  plate  currents  of  the  output  tubes  is 
cancelled  in  the  windings  of  the  output  trans- 
former; (2)  even  harmonics  of  the  input  signal 
(second  and  fourth  harmonics  primarily)  gen- 
erated in  the  push-pull  stage  are  cancelled 
when  the  tubes  are  balanced. 

The  cancellation  of  even  harmonics  gener- 
ated in  the  stage  allows  the  tubes  to  be  oper- 


0 + B PLATE 
PUSH-PULL  TRIODE  AND  TETRODE 


0+BS.g.  0+B  plate 


figure  21 


122  Vacuum  Tube  Amplifiers 


THE  RADIO 


® ® 


Figure  22 

DETERMINATION  OF  OPERATING  PARAMETERS  FOR  PUSH-PULL  CLASS  A 

TRIODE  TUBES 


ated  Class  AB  — in  other  words  the  tubes  may 
be  operated  with  bias  and  input  signals  of 
such  amplitude  that  the  plate  current  of  alter- 
nate tubes  may  be  cut  off  during  a portion  of 
the  input  voltage  cycle.  If  a tube  were  operated 
in  such  a manner  in  a single-ended  amplifier 
the  second  harmonic  amplitude  generated  would 
be  prohibitively  high. 

. Push-pull  Class  AB  operation  allows  a plate 
circuit  efficiency  of  from  45  to  60  per  cent  to 
be  obtained  in  an  amplifier  stage  depending 
upon  whether  or  not  the  exciting  voltage  is 
of  such  amplitude  that  grid  current  is  drawn 
by  the  tubes.  If  grid  current  is  taken  on  input 
voltage  peaks  the  amplifier  is  said  to  be  oper- 
ating Class  AB2  and  the  plate  circuit  effi- 
ciency can  be  as  high  as  the  upper  value  just 
mentioned.  If  grid  currenr  is  nor  taken  by  the 
stage  it  is  said  to  be  operating  Class  ABj  and 
the  plate  circuit  efficiency  will  be  toward  rhe 
lower  end  of  the  range  just  quoted.  In  all  Class 
AB  amplifiers  the  plate  current  will  increase 
from  40  to  150  per  cent  over  the  no-signal 
value  when  full  signal  is  applied. 


Operating  Characteristics  The  operating  char- 
of  Push-Pull  Class  A acteristics  of  push- 

Triade  Rawer  Amplifier  pull  Class  A ampli- 
fiers may  also  be 
determined  from  the  plate  family  of  curves  for 


a particular  triode  tube  by  the  following  steps; 

1-  Erect  a vertical  line  from  the  plate  volt- 
age axis  (x-axis)  at  0.  6 Ep  (figure  22), 
which  intersects  the  Eg  = 0 curve.  This 
point  of  intersection  (P),  interpolated  to 
the  plate  current  axis  (y-axis)  may  be 
taken  as  inax-  It  is  assumed  for  simplifi- 
cation that  inax  occurs  at  the  point  of 
the  zero-bias  curve  corresponding  to 
0.6  Ep. 

2-  The  power  output  obtainable  from  the  two 
tubes  is: 

Y-i  /r»  \ ^max  ^ Ep 

Power  output  (Pp)  = 

where  Pq  is  expressed  in  watts,  in,ax  in 
amperes,  and  Ep  is  the  applied  plate 
voltage. 

3-  Draw  a preliminary  load  line  through 
point  P to  the  Ep  point  located  on  the 
x-axis  (the  zero  plate  current  line).  This 
load  line  represents  % of  the  actual  plate- 
to-plate  load  of  the  Class  A tubes.  There- 
fore: 

E - 0.6  Ep 

Rl  (plate-to-plate)  = 4 x — — 

^max 

1-6  Ep 
^max 


HANDBOOK 


Class  B Audio  Amplifiers 


123 


where  Rl  is  expressed  in  ohms,  Ep  in 
volts,  and  in  amperes. 

Figure  22  illustrates  the  above  steps  ap- 
plied to  a push-pull  Class  A amplifier  using 
two  2A3  tubes. 

4-  The  average  plate  current  is  0.636  imau 
and,  multiplied  by  the  plate  voltage,  Ep, 
will  give  the  average  watts  input  to  the 
plates  of  the  two  tubes.  The  power  out- 
put should  be  subtracted  from  this  value 
to  obtain  the  total  operating  plate  dis- 
sipation of  the  two  tubes.  If  the  plate 
dissipation  is  excessive,  a slightly 
higher  value  of  should  be  chosen  to 
limit  the  plate  dissipation. 

5-  The  correct  value  of  operating  bias,  and 
the  static  plate  current  for  the  push-pull 
tubes  may  be  determined  from  the  Eg  vs. 
Ip  curves,  which  are  a derivation  of  the 
Ep  vs.  Ip  curves  for  various  values  of 
Eg. 

6-  The  Eg  vs.  Ip  curve  may  be  constructed 
in  this  manner;  Values  of  grid  bias  are 
read  from  the  intersection  of  each  grid 
bias  curve  with  the  load  line.  These 
points  are  transferred  to  the  Eg  vs.  Ip 
graph  to  ptoduce  a curved  line,  A-B.  If 
the  grid  bias  curves  of  the  Ep  vs.  Ip 
gtaph  were  straight  lines,  the  lines  of 
the  Eg  vs.  Ip  graph  would  also  be  straight 
This  is  usually  not  the  case.  A tangent 
to  this  curve  is  therefore  drawn,  starting 
at  point  A ',  and  intersecting  the  grid 
voltage  abscissa  (x-axis).  This  inter- 
section (C)  is  the  operating  bias  point 
for  fixed  bias  operation. 

7-  This  operating  bias  point  may  now  be 
plotted  on  the  original  Eg  vs.  Ip  family 
of  curves  (C'),  and  the  zero-signal  cur- 
rent produced  by  this  bias  is  determined. 

This  operating  bias  point  (C  ')  does  not  fall 
on  the  operating  load  line,  as  in  the  case  of  a 
single-ended  amplifier. 

8-  Under  conditions  of  maximum  power  out- 
put, the  exciting  signal  voltage  swings 
from  zero-bias  voltage  to  zero-bias  volt- 
age for  each  of  the  tubes  on  each  half 
of  the  signal  cycle.  Second  harmonic 
distortion  is  largely  cancelled  out. 


6-13  Class  B Audio  Frequency 
Power  Amplifiers 

The  Class  B audio-frequency  power  ampli- 
fier (figure  23)  operates  at  a higher  plate- 

circuit  efficiency  than  any  of  the  previously 

described  types  of  audio  power  amplifiers. 
Full-signal  plate-circuit  efficiencies  of  60  to 


OPERATING 

CONDITION) 

Figure  23 

CLASS  B AUDIO  FREQUENCY 
POWER  AMPLIFIER 


70  pet  cent  ate  readily  obtainable  with  the 
tube  types  at  present  available  for  this  type 
of  work.  Since  the  plate  circuit  efficiency  is 
higher,  smaller  tubes  of  lower  plate  dissipa- 
tion may  be  used  in  a Class  B power  amplifier 
of  a given  power  output  than  can  be  used  in 
any  other  conventional  type  of  audio  amplifier. 
An  additional  factor  in  favor  of  the  Class  B 
audio  amplifier  is  the  fact  that  the  power  in- 
put to  the  stage  is  relatively  low  under  no- 
signal conditions.  It  is  for  these  reasons  that 
this  type  of  amplifier  has  largely  superseded 
other  types  in  the  generation  of  audio-frequency 
levels  from  perhaps  100  watts  on  up  to  levels 
of  approximately  150,000  watts  as  required  for 
large  short-wave  broadcast  stations. 

Disadvantages  of  There  are  attendant  dis- 
Class  B Amplifier  advantageous  features  to  the 
Operation  operation  of  a power  ampli- 

fier of  this  type;  but  all 
these  disadvantages  can  be  overcome  by  proper 
design  of  the  circuits  associated  with  the 
power  amplifier  stage.  These  disadvantages 
are:  (1)  The  Class  B audio  amplifier  requires 
driving  power  in  its  grid  circuit;  this  dis- 
advantage can  be  overcome  by  the  use  of  an 
oversize  power  stage  preceding  the  Class  B 
stage  with  a step-down  transformet  between 
the  driver  stage  and  the  Class-B  grids.  De- 
generative feedback  is  sometimes  employed 
to  reduce  the  plate  impedance  of  the  driver 
stage  and  thus  to  improve  the  voltage  regula- 
tion under  the  varying  load  presented  by  the 
Class  B grids.  (2)  The  Class  B stage  requires 
a constant  value  of  average  grid  bias  to  be 
supplied  in  spite  of  the  fact  that  the  grid  cur- 
rent on  the  stage  is  zero  over  most  of  the 
cycle  but  rises  to  values  as  high  as  one-third 
of  the  peak  plate  current  on  the  peak  of  the 
exciting  voltage  cycle.  Special  regulated  bias 
supplies  have  been  used  for  this  application, 
or  B batteries  can  be  used.  However,  a number 


124  Vacuum  Tube  Amplifiers 


THE  RADIO 


of  tubes  especially  designed  for  Class  B audio 
amplifiers  have  been  developed  which  require 
zero  average  grid  bias  for  their  operation.  The 
811A,  838,  805,  809,  HY-5514,  and  TZ-40  are 
examples  of  this  type  of  tube.  All  these  so- 
called  "zero-bias”  tubes  have  rated  operating 
conditions  up  to  moderate  plate  voltages 
wherein  they  can  be  operated  without  grid 
bias.  As  the  plate  voltage  is  increased  to 
to  their  maximum  ratings,  however,  a small 
amount  of  grid  bias,  such  as  could  be  obtained 
from  several  4 Vi-voXi  C batteries,  is  required. 

(3),  A Class  B audio-frequency  power  ampli- 
fier or  modulator  requires  a source  of  plate 
supply  voltage  having  reasonably  good  regula- 
tion. This  requirement  led  to  the  development 
of  the  swinging  choke.  The  swinging  choke  is 
essentially  a conventional  filter  choke  in 
which  the  cote  air  gap  has  been  reduced.  This 
reduction  in  the  ait  gap  allows  the  choke  to 
have  a much  greater  value  of  inductance  with 
low  current  values  such  as  are  encountered 
with  no  signal  or  small  signal  being  applied 
to  the  Class  B stage.  With  a higher  value  of 
current  such  as  would  be  taken  by  a Class  B 
stage  with  full  signal  applied  the  inductance 
of  the  choke  drops  to  a much  lower  value. 
With  a swinging  choke  of  this  type,  having 
adequate  current  rating,  as  the  input  inductor 
in  the  filter  system  for  a rectifier  power  sup- 
ply, the  regulation  will  be  improved  to  a point 
which  is  normally  adequate  for  a power  supply 
for  a Class  B amplifier  or  modulator  stage. 

Calculation  of  Operating  The  following  proce- 
Conditions  of  Class  B dure  can  be  used  for 
Power  Amplifiers  the  calculation  of  the 

operating  conditions 
of  Class  B power  amplifiers  when  they  ate  to 
operate  into  a resistive  load  such  as  the  type 
of  load  presented  by  a Class  C power  ampli- 
fier. This  procedure  will  be  found  quite  satis- 
factory for  the  application  of  vacuum  tubes  as 
Class  B modulators  when  it  is  desired  to 
operate  the  tubes  under  conditions  which  are 
not  specified  in  the  tube  operating  character- 
istics published  by  the  tube  manufacturer.  The 
same  procedure  can  be  used  with  equal  effec- 
tiveness for  the  calculation  of  the  operating 
conditions  of  beam  tetrodes  as  Class  AB2 
amplifiers  or  modulators  when  the  resting 
plate  current  on  the  tubes  (no  signal  condi- 
tion) is  less  than  25  or  30  per  cent  of  the 
maximum-signal  plate  current. 

1-  With  the  average  plate  characteristics 
of  the  tube  as  published  by  the  manu- 
facturer before  you,  select  a point  on 
the  Ep  = Eg  (diode  bend)  line  at  about 
twice  the  plate  current  you  expect  the 
tubes  to  kick  to  under  modulation.  If 
beam  tetrode  tubes  are  concerned,  select 


a point  at  about  the  same  amount  of  plate 
current  mentioned  above,  just  to  the 
right  of  the  region  where  the  Ij,  line 
takes  a sharp  curve  downward.  This  will 
be  the  first  trial  point,  and  the  plate 
voltage  at  the  point  chosen  should  be 
not  more  than  about  20  pet  cent  of  the 
d-c  voltage  applied  to  the  tubes  if  good 
plate-circuit  efficiency  is  desired. 

2-  Note  down  the  value  of  ipmax  ^nd  epnjn  at 
this  point. 

3-  Subtract  the  value  of  Cp^^jn  from  the  d-c 
plate  voltage  on  the  tubes. 

4-  Substitute  the  values  obtained  in  the 
following  equations: 

ipmax  (Ebb  - ^pmin) 

Pq  = — = Power  output 

2 from  2 tubes 

.j.  , (^bb  ~ ^poiin) 

Rl  = 4 ; 

^pmax 

= Plate-to-plate  load  for  2 tubes 
Full  signal  efficiency  (Np)  « 


Effects  of  Speech  All  the  above  equations  are 
Clipping  true  for  sine-wave  operating 

conditions  of  the  tubes  con- 
cerned. However,  if  a speech  clipper  is  being 
used  in  the  speech  amplifier,  or  if  it  is  desired 
to  calculate  the  operating  conditions  on  the 
basis  of  the  fact  that  the  ratio  of  peak  power 
to  average  power  in  a speech  wave  is  approxi- 
mately 4-to-l  as  contrasted  to  the  ratio  of 
2-to-l  in  a sine  wave— -in  other  words,  when 
non-sinusoidal  waves  such  as  plain  speech  or 
speech  that  has  passed  through  a clipper  are 
concerned,  we  are  no  longer  concerned  with 
average  power  output  of  the  modulator  as  far 
as  its  capability  of  modulating  a Class-C  ampli- 
fier is  concerned;  we  are  concerned  with  its 
peak‘powefoutput  capability. 

Under  these  conditions  we  call  upon  other, 
more  general  relationships.  The  first  of  these 
is:  It  requires  a peak  power  output  equal  to 
the  Class-C  stage  input  to  modulate  that  input 
fully. 

The  second  one  is:  The  average  power  out- 
put required  of  the  modulator  is  equal  to  the 
shape  factor  of  the  modulating  wave  multi- 
plied by  the  input  to  the  Class-C  stage.  The 
shape  factor  of  undipped  speech  is  approxi- 
mately 0.  25*  The  shape  factor  of  a sine  wave 
is  0.  5-  The  shape  factor  of  a speech  wave  that 


HANDBOOK 


Class  B Parameters  125 


Figure  24 

Typical  Class  B a~f  amplifier 
load  line.  The  load  line  has 
been  drawn  on  the  average 
characteristics  of  a type  811 
tube. 


has  been  passed  through  a clipper-filter  ar- 
rangement is  somewhere  between  0.  25  and  0.  9 
depending  upon  the  amount  of  clipping  that 
has  taken  place.  With  15  or  20  db  of  clipping 
the  shape  factor  may  be  as  high  as  the  figure 
of  0.  9 mentioned  above.  This  means  that  the 
audio  power  output  of  the  modulator  will  be 
90%  of  the  input  to  the  Class-C  stage.  Thus 
with  a kilowatt  input  we  would  be  putting 
900  watts  of  audio  into  the  Class-C  stage  for 
100  per  cent  modulation  as  contrasted  to  per- 
haps 250  watts  for  undipped  speech  modula- 
tion of  100  per  cent. 

Sample  Calculation  Figure  24  shows  a set  of 
for  811 A Tubes  plate  characteristics  for 
a type  811A  tube  with  a 
load  line  for  Class  B operation.  Figure  25 
lists  a sample  calculation  for  determining  the 
proper  operating  conditions  for  obtaining  ap- 
proximately 185  watts  output  from  a pair  of 
the  tubes  with  1000  volts  d-c  plate  potential. 
Also  shown  in  figure  25  is  the  method  of  de- 
termining the  proper  ratio  for  the  modulation 
transformer  to  couple  between  the  811’s  or 
SllA’s  and  the  anticipated  final  amplifier 
which  is  to  operate  at  2000  plate  volts  and 
175  ma.  plate  current. 

Modulation  Transformer  The  method  illustrated 
Colculotion  in  figure  25  can  be  used 

in  general  for  the  deter- 
mination of  the  proper  transformer  ratio  to 
couple  between  the  modulator  tube  and  the 
amplifier  to  be  modulated.  The  procedure  can 

be  Stated  as  follows:  (1)  Determine  the  proper 

plate-to-plate  load  impedance  for  the  modulator 
tubes  either  by  the  use  of  the  type  of  calcula- 


tion shown  in  figure  25,  or  by  reference  to  the 
published  characteristics  on  the  tubes  to  be 
used.  (2)  Determine  the  load  impedance  which 
will  be  presented  by  the  Class  C amplifier 
stage  to  be  modulated  by  dividing  the  operating 
plate  voltage  on  that  stage  by  the  operating 
value  of  plate  current  in  amperes.  (3)  Divide 
the  Class  C load  impedance  determined  in  (2) 


SAMPLE  CALCULATION 

CONDITION:  2 TYPE  811  TUBES,  Ebb,  = 1000 
INPUT  TO  riNAL  STAGE,  350  W. 

PEAK  POWER  OUTPUT  NEE0E0=  350  + 6«lb  = 370  W, 
FINAL  AMPLIFIER  Ebb  * 2000  V. 

FINAL  AMPLIFIER  Lb  = .175  A. 

FINAL  AMPLIFIER  ZL  = 2000  = IIAOOA 

.175 

EXAMPLE:  CHOSE  POINT  ON  811  CHARACTERISTICS  JUST 
TO  RIGHTOF  Ebb=EcC.  ^ ^f*fr  F/^.  2.^  ) 
Ip  MAX.  = .A10  A.  Ep  MIN,  = +100 

IG  MAX.  = . 100  A.  Eg  MAX.  = + 80 

PEAK  Po  = .410  X (lOOO-lOO)  = .410  X 900  = 369  W. 

RL  = 4 X -1^  = 8800  a. 

Np  = 76.5  tl-:f^)=  76.5  (.9)=  70.5  1b 
Wo  (average  with  SINE  WAVE)  = 

Win  = ^■^*1  = 260  w. 

Ib  (maximum  with  SINE  wave)  = 260  MA 
Wg  peak  = . 100  X 60  = 6W. 

DRIVING  POWER  = = 4 W. 

TRANSFORMER: 


TURNS  RATIO  = OiT  = ^ 1.29  =1,14  STEP  UP 

Zp 

Figure  25 

Typical  calculation  of  operating  conditions  for 
0 Closs  B o-f  power  ampliflor  using  a pair  of 
type  811  or  811A  tubes.  Plate  c/iarocfer/sr/cs 
and  load  line  shown  in  figure  24. 


126  Vacuum  Tube  Amplifiers 


THE  RADIO 


above  by  the  plate-to-plate  load  impedance  for 

the  modulator  tubes  determined  in  (1)  above. 
The  ratio  determined  in  this  way  is  the  sec- 
ondary-to-primary  impedance  ratio.  (4)  Take 
the  square  root  of  this  ratio  to  determine  the 
secondary-to-primary  turns  ratio.  If  the  turns 
ratio  is  greater  than  one  the  use  of  a step-up 
transformer  is  required.  If  the  turns  ratio  as 
determined  in  this  way  is  less  than  one  a step- 
down  transformer  is  called  for. 

If  the  procedure  shown  in  figure  ‘25  has 
been  used  to  calculate  the  operating  conditions 
for  the  modulator  tubes,  the  transformer  ratio 
calculation  can  be  checked  in  the  following 
manner;  Divide  the  plate  voltage  on  the  mod- 
ulated amplifier  by  the  total  voltage  swing  on 
the  modulator  tubes:  2 (Ebb  ~ Cmin)-  This  ratio 
should  be  quite  close  numerically  to  the  trans- 
former turns  ratio  as  previously  determined. 
The  reason  for  this  condition  is  that  the  ratio 
between  the  total  primary  voltage  and  the  d-c 
plate  supply  voltage  on  the  modulated  stage 
is  equal  to  the  turns  ratio  of  the  transformer, 
since  a peak  secondary  voltage  equal  to  the 
plate  voltage  on  the  modulated  stage  is  re- 
quired to  modulate  this  stage  100  per  cent. 

Use  of  Clipper  Speech  When  a clipper  speech 
Amplifier  with  Tetrode  amplifier  is  used  in 
Modulator  Tubes  conjunction  with  a Class 

B modulator  stage,  the 
plate  current  on  that  stage  will  kick  to  a 
higher  value  with  moduiation(due  to  the  greater 
average  power  output  and  input)  but  the  plate 
dissipation  on  the  tubes  will  ordinarily  be 
less  than  with  sine-wave  modulation.  However, 
when  tetrode  tubes  ate  used  as  modulators, 
the  screen  dissipation  will  be  much  greater 
than  with  sine-wave  modulation.  Care  must 
be  taken  to  insure  that  the  screen  dissipa- 
tion rating  on  the  modulator  tubes  is  not  ex- 
ceeded under  full  modulation  conditions  with 
a clipper  speech  amplifier.  The  screen  dissipa- 
tion is  equal  to  screen  voltage  times  screen 
current. 

Practical  Aspects  of  As  stated  previously,  a 
Class  B Modulators  Class  B audio  amplifier 
requires  the  driving  stage 
to  supply  well-regulated  audio  power  to  the 
grid  circuit  of  the  Class  B stage.  Since  the 
performance  of  a Class  B modulator  may  easily 
be  impaired  by  an  improperly  designed  driver 
stage,  it  is  well  to  study  the  problems  in- 
curred in  the  design  of  the  driver  stage. 

The  grid  circuit  of  a Class  B modulator  may 
be  compared  to  a variable  resistance  which 
decreases  in  value  as  the  exciting  grid  volt- 
age is  increased.  This  variable  resistance  ap- 
pears across  the  secondary  terminals  of  the 
driver  transformer  so  that  the  driver  stage  is 


called  upon  to  deliver  power  to  a varying  load. 
For  best  operation  of  the  Class  B stage,  the 
grid  excitation  voltage  should  not  drop  as  the 
power  taken  by  the  grid  circuit  increases. 
These  opposing  conditions  call  for  a high  or- 
der of  voltage  regulation  in  the  driver  stage 
plate  circuit.  In  order  to  enhance  the  voltage 
regulation  of  this  circuit,  the  driver  tubes  must 
have  low  plate  resistance,  the  driver  trans- 
former must  have  as  large  a step-down  ratio 
as  possible,  and  the  d-c  resistance  of  both 
primary  and  secondary  windings  of  the  driver 
transformer  should  be  low. 

The  driver  transformer  should  reflect  into 
the  plate  circuit  of  the  driver  tubes  a load  of 
such  value  that  the  required  driving  power  is 
just  developed  with  full  excitation  applied  to 
the  driver  grid  circuit.  If  this  is  done,  the 
driver  transformer  will  have  as  high  a step- 
down  ratio  as  is  consistent  with  the  maximum 
drive  requirements  of  the  Class  B stage.  If 
the  step-down  ratio  of  the  driver  transformer  is 
too  large,  the  driver  plate  load  will  be  so 
high  that  the  power  required  to  drive  the  Class 
B stage  to  full  output  cannot  be  developed. 
If  the  step-down  ratio  is  too  small  the  regula- 
tion of  the  driver  stage  will  be  impaired. 


Driver  Stage  The  parameters  for  the  driver 
Calculations  stage  may  be  calculated  from 
the  plate  characteristic  curve,  a 
sample  of  which  is  shown  in  figure  24.  The 
required  positive  grid  voltage  (eg.^a,)  for  the 
811A  tubes  used  in  the  sample  calculation  is 
found  at  point  X,  the  intersection  of  the  load 
line  and  the  peak  plate  current  as  found  on  the 
y-axis.  This  is  + 80  volts.  If  a vertical  line 
is  dropped  from  point  X to  intersect  the  dotted 
grid  current  curves,  it  will  be  found  that  the 
grid  current  for  a single  81  lA  at  this  value  of 
grid  voltage  is  100  milliamperes  (point  Y). 
The  peak  grid  driving  power  is  therefore 
80  X 0. 100  = 8 watts.  The  approximate  average 
driving  power  is  4 watts.  This  is  an  approxi- 
mate figure  because  the  grid  impedance  is  not 
constant  over  the  entire  audio  cycle. 

A pair  of  2A3  tubes  will  be  used  as  drivers, 
(derating  Class  A,  with  the  maximum  excita- 
tion to  the  drivers  occuring  just  below  the 
point  of  grid  current  flow  in  the  2A3  tubes. 
The  driver  plate  voltage  is  300  volts,  and  the 
grid  bias  is  —62  volts.  The  peak  power  devel- 
oped in  the  primary  winding  of  the  driver 
transformer  is: 


Peak  Power  (Pp) 
(watts) 


2Rl 


/ 

Wp  +Rl 


) 


a 


where  p is  the  amplification  factor  of  the 
driver  tubes  (4.2  for  2A3).  Eg  is  the  peak  grid 
swing  of  the  driver  stage  (62  volts).  Rp  is  the 


HANDBOOK 


Cathode  Follower  Amplifier  127 


plate  resistance  of  one  driver  tube  (800  ohms). 
Rl  Vi-  fhe  plate-to-plate  load  of  the  driver 
stage,  and  Pp  is  8 watts. 

Solving  the  above  equation  for  Rl,  we 
obtain  a value  of  14,500  ohms  load,  plate-to- 
plate  for  the  2A3  driver  tubes. 

The  peak  primary  voltage  is: 

ep,i  = 2Rl  X = 493  volts 

Rp  + Rl 

and  the  turns  ratio  of  the  driver  transformer 
(primary  to  secondary)  is: 


^g(max) 

Plate  Circuit  One  of  the  commonest  causes  of 
Impedance  distortion  in  a Class  B modu- 
Matching  lator  is  incorrect  load  impedance 
in  the  plate  circuit.  The  purpose 
of  the  Class  B modulation  transformer  is  to 
take  the  power  developed  by  the  modulator 
(which  has  a certain  operating  impedance)  and 
transform  it  to  the  operating  impedance  im- 
posed by  the  modulated  amplifier  stage. 

If  the  transformer  in  question  has  the  same 
number  of  turns  on  the  primary  winding  as  it 
has  on  the  secondary  winding,  the  turns  ratio 
is  1:1,  and  the  impedance  ratio  is  also  1:1.  If 
a 10,  000  ohm  resistor  is  placed  across  the 
secondary  terminals  of  the  transformer,  a re- 
flected load  of  10, 000  ohms  would  appear 
across  the  primary  terminals.  If  the  resistor 
is  changed  to  one  of  2376  ohms,  the  reflected 
primary  impedance  would  also  be  2376  ohms. 

If  the  transformer  has  twice  as  many  turns 
on  the  secondary  as  on  the  primary,  the  turns 
ratio  is  2:1.  The  impedance  ratio  is  the  square 
of  the  turns  ratio,  or  4:1.  If  a 10,000  ohm 
resistor  is  now  placed  across  the  secondary 
winding,  a reflected  load  of  2,500  ohms  will 
appear  across  the  primary  winding. 

Effects  of  Plate  It  can  be  seen  from  the 
Circuit  Mis-match  above  paragraphs  that  the 
Class  B modulator  plate 
load  is  entirely  dependent  upon  the  load 
placed  upon  the  secondary  terminals  of  the 
Class  B modulation  transformer.  If  the  second- 
ary load  is  incorrect,  certain  changes  will 
take  place  in  the  operation  of  the  Class  B 
modulator  stage. 

When  the  modulator  load  impedance  is  too 
low,  the  efficiency  of  the  Class  B stage 
is  reduced  and  the  plate  dissipation  of  the 
tubes  is  increased.  Peak  plate  current  of 
the  modulator  stage  is  increased,  and  satura- 
tion of  the  modulation  transformer  core  may 
result.  "Talk-back”  of  the  modulation  trans- 


former may  result  if  the  plate  load  impedance 
of  the  modulator  stage  is  too  low. 

When  the  modulator  load  impedance  is  too 
high,  the  maximum  power  capability  of  the 
stage  is  reduced.  An  attempt  to  increase  the 
output  by  increasing  grid  excitation  to  the 
stage  will  result  in  peak-clipping  of  the  audio 
wave.  In  addition,  high  peak  voltages  may  be 
built  up  in  the  plate  circuit  that  may  damage 
the  modulation  transformer. 


6-14  Cathode-Follower 

Power  Amplifiers 

The  cathode-follower  is  essentially  a power 
output  stage  in  which  the  exciting  signal  is 
applied  between  grid  and  ground.  The  plate  is 
maintained  at  ground  potential  with  respect  to 
input  and  output  signals,  and  the  output  signal 
is  taken  between  cathode  and  ground. 

Types  of  Cathode-  Figure  26  illustrates  four 
Follower  Amplifiers  types  of  cathode-follower 
power  amplifiers  in  com- 
mon usage  and  figure  27  shows  the  output 
impedance  (Ro),  and  stage  gain  (A)  of  both 
triode  and  pentode  (or  tetrode)  cathode-follower 
stages.  It  will  be  seen  by  inspection  of  the 
equations  that  the  stage  voltage  gain  is  always 
less  than  one,  that  the  output  impedance  of 
the  stage  is  much  less  than  the  same  stage 
operated  as  a conventional  cathode-return 
amplifier.  The  output  impedance  for  con- 
ventional tubes  will  be  somewhere  between 
100  and  1000  ohms,  depending  primarily  on 
the  transconductance  of  the  tube. 

This  reduction  in  gain  and  output  imped- 
ance for  the  cathode-follower  comes  about 
since  the  stage  operates  as  though  it  has  100 
per  cent  degenerative  feedback  applied  between 
its  output  and  input  circuit.  Even  though  the 
voltage  gain  of  the  stage  is  reduced  to  a value 
less  than  one  by  the  action  of  the  degenerative 
feedback,  the  power  gain  of  the  stage  (if  it  is 
operating  Class  A)  is  not  reduced.  Although 
mote  voltage  is  required  to  excite  a cathode- 
follower  amplifier  than  appears  across  the  load 
circuit,  since  the  cathode  "follows”  along 
with  the  grid,  the  relative  grid-to-cathode  volt- 
age is  essentially  the  same  as  in  a con- 
ventional amplifier. 

Use  of  Cathode-  Although  the  cathode-fol- 
Follower  Amplifiers  lower  gives  no  voltage 
gain,  it  is  an  effective 
power  amplifier  where  it  is  desired  to  feed  a 
low-impedance  load,  or  where  it  is  desired  to 
feed  a load  of  varying  impedance  with  a signal 
having  good  regulation.  This  latter  capability 


128  Vacuum  Tube  Amplifiers 


THE  RADIO 


Figure  26 

CATHODE-FOLLOWER  OUTPUT 
CIRCUITS  FOR  AUDIO  OR 
VIDEO  AMPLIFIERS 


makes  the  cathode  follower  particularly  effec- 
tive as  a driver  for  the  grids  of  a Class  B 
modulator  stage. 

The  circuit  of  figure  26A  is  the  type  of  am- 
plifier, either  single-ended  or  push-pull,  which 
may  be  used  as  a driver  for  a Class  B modu- 
lator or  which  may  be  used  for  other  applica- 
tions such  as  feeding  a loudspeaker  where  un- 
usually good  damping  of  the  speaker  is  de- 
sired. If  the  d-c  resistance  of  the  primary  of 
the  transformer  T2is  approximately  the  correct 
value  for  the  cathode  bias  resistor  for  the  am- 


TRIOOE;  ,,  JJ 

JUcF  = 

Xl  + 1 


Ro(cathooe)  = 


±£- 
Ai  + 1 


^ = ■ jg-R  -L  _ 

Rl(jU+  0 + Rp 

Rl  = + Rk2)  Rl^ 

RkI  + Rk2+  Rl' 


PENTODE. 


Gm 


A = Gm  Rbq 


Reg  = 


Rl 

1 + R (_  Gm 


Figure  27 

Equivalent  foctors  for  pentode  (or  tetrode) 
cathode-foUower  power  amplifiers. 


plifier  tube,  the  components  Rj;  and  C);  need 
not  be  used.  Figure  26B  shows  an  arrangement 
which  may  be  used  to  feed  directly  a value  of 
load  impedance  which  is  equal  to  or  higher 
than  the  cathode  impedance  of  the  amplifier 
tube.  The  value  of  C^  must  be  quite  high, 
somewhat  higher  than  would  be  used  in  a con- 
ventional circuit,  if  the  frequency  response  of 
the  circuit  when  operating  into  a low-imped- 
ance load  is  to  be  preserved. 

Figures  26C  and  26D  show  cathode-follower 
circuits  for  use  with  tetrode  or  pentode  tubes. 
Figure  26C  is  a circuit  similar  to  that  shown 
in  26A  and  essentially  the  same  comments 
apply  in  regard  to  the  components  Ri,  and 
and  the  primary  resistance  of  the  transfotmer 
T2.  Notice  also  that  the  screen  of  the  tube  is 
maintained  at  the  same  signal  potential  as  the 
cathode  by  means  of  coupling  capacitor  Cj. 
This  capacitance  should  be  large  enough  so 
that  at  the  lowest  frequency  it  is  desired  to 
pass  through  the  stage  its  reactance  will  be 
low  with  respect  to  the  dynamic  screen-to- 
cathode  resistance  in  parallel  with  Rj  T2  in 
this  stage  as  well  as  in  the  circuit  of  figure 
26A  should  have  the  proper  turns  (or  imped- 
ance) ratio  to  give  the  desired  step-down  or 
step-up  from  the  cathode  circuit  to  the  load. 
Figure  26D  is  an  arrangement  frequently  used 
in  video  systems  for  feeding  a coaxial  cable  of 
relatively  low  impedance  from  a vacuum-tube 
amplifier.  A pentode  or  tetrode  tube  with  a 
cathode  impedance  as  a cathode  follower 
(1/Gn,)  approximately  the  same  as  the  cable 
impedance  should  be  chosen.  The  6AG7  and 
6AC7  have  cathode  impedances  of  the  same 
order  as  the  surge  impedances  of  certain  types 
of  low-capacitance  coaxial  cable.  An  arrange- 
ment such  as  26D  is  also  usable  for  feeding 
coaxial  cable  with  audio  or  r-f  energy  where 
it  is  desired  to  transmit  the  output  signal 
over  moderate  distances.  The  resistor  Rj;  is 
added  to  the  citcuit  as  shown  if  the  cathode 
impedance  of  the  tube  used  is  lower  than  the 


HANDBOOK 


Feedback  Amplifiers 


129 


characteristic  impedance  of  the  cable.  If  the 
output  impedance  of  the  stage  is  higher  than 
the  cable  impedance  a resistance  of  appro- 
priate value  is  sometimes  placed  in  parallel 
with  the  input  end  of  the  cable.  The  values 
of  Cj  and  Rj  should  be  chosen  with  the  same 
considerations  in  mind  as  mentioned  in  the 
discussion  of  the  circuit  of  figure  26C  above. 

The  Cathode-Follower  The  cathode  follower 
in  R-F  Sfoges  may  conveniently  be 

used  as  a method  of  cou- 
pling r-f  or  i-f  energy  between  two  units  sepa- 
rated a considerable  distance.  In  such  an 
application  a coaxial  cable  should  be  used  to 
carry  the  r-f  or  i-f  energy.  One  such  applica- 
tion would  be  for  carrying  the  output  of  a v-f-o 
to  a transmitter  located  a considerable  dis- 
tance from  the  operating  position.  Another 
application  would  be  where  it  is  desired  to 
feed  a single-sideband  demodulator,  an  FM 
adaptor,  or  another  accessory  with  inter- 
mediate frequency  signal  from  a communica- 
tions receiver.  A tube  such  as  a 6CB6  con- 
nected in  a manner  such  as  is  shown  in  figure 
26D  would  be  adequate  for  the  i-f  amplifier 
coupler,  while  a 6L6  or  a 6AG7  could  be  used 
in  the  output  stage  of  a v-f-o  as  a cathode 
follower  to  feed  the  coaxial  line  which  carries 
the  v-f-o  signal  from  the  control  unit  to  the 
transmitter  proper. 


6-15  Feedback  Amplifiers 


It  is  possible  to  modify  the  characteristics 
of  an  amplifier  by  feeding  back  a portion  of 
the  output  to  the  input.  All  components,  cir- 
cuits and  tubes  included  between  the  point 
where  the  feedback  is  taken  off  and  the  point 
where  the  feedback  energy  is  inserted  are 
said  to  be  included  within  the  feedback  loop. 
An  amplifier  containing  a feedback  loop  is 
said  to  be  a feedback  amplifier.  One  stage  or 
any  number  of  stages  may  be  included  within 
the  feedback  loop.  However,  the  difficulty  of 
obtaining  proper  operation  of  a feedback  am- 
plifier increases  with  the  bandwidth  of  the 
amplifier,  and  with  the  number  of  stages  and 
circuit  elements  included  within  the  feedback 
loop. 

Gain  and  Phase-shift  The  gain  and  phase 
in  Feedback  Amplifiers  shift  of  any  amplifier 
are  functions  of  fre- 
quency. For  any  amplifier  containing  a feed- 
back loop  to  be  completely  stable  the  gain  of 
such  an  amplifier,  as  measured  from  the  input 
back  to  the  point  where  the  feedback  circuit 
connects  to  the  input,  must  be  less  than  one 


INPUT  SIGNAL 


Ec  AMPLIFICR 

H eAIN  = A 


FEEDBACK  OR  B PATH 


OUTPUT  Eq 


VOLTAGE  AMPLIFICATION  WITH  FEEDBACK^  

1-A  6 

A = GAIN  IN  ABSENCE  OF  FEEDBACK 


B = FRACTION  OF  OUTPUT  VOLTAGE  FED  BACK 


6 IS  NEGATIVE  FOR  NEGATIVE  FEEDBACK 
FEEDBACK  IN  DECIBELS  = 20  LOG  (l  -A  ) 


MID  FREQ.  CAIN  WITHOUT  FEEDBACK 
MID  FREQ,  GAIN  WITH  FEEDBACK 


DISTORTION  WITH  FEEDBACK  s 


DISTORTION  WITHOUT  FEEDBACK 
(1-Aa) 


WHERE- 


Ro  = 


-Afl  {l 


Ro  = OUTPUT  IMPEDANCE  OF  AMPLIFIER  WITH  FEEDBACK 
Rn  = OUTPUT  IMPEDANCE  OF  AMPLIFIER  WITHOUT  FEEDBACK 
Rt  s LOAD  IMPEDANCE  INTO  WHICH  AMPLIFIER  OPERATES 


Figure  28 

FEEDBACK  AMPLIFIER  RELATIONSHIPS 


at  the  frequency  where  the  feedback  voltage 
is  in  phase  with  the  input  voltage  of  the  am- 
plifier. If  the  gain  is  equal  to  or  more  than 
one  at  the  frequency  where  the  feedback  volt- 
age is  in  phase  with  the  input  the  amplifier 
will  oscillate.  This  fact  imposes  a limitation 
upon  the  amount  of  feedback  which  may  be 
employed  in  an  amplifier  which  is  to  remain 
stable.  If  the  reader  is  desirous  of  designing 
amplifiers  in  which  a large  amount  of  feed- 
back is  to  be  employed  he  is  referred  to  a 
book  on  the  subject  by  H.  W.  Bode.* 

Types  of  Feedback  may  be  either  negative 
Feedback  or  positive,  and  the  feedback  volt- 
age may  be  proportional  either  to 
output  voltage  or  output  current.  The  most 
commonly  used  type  of  feedback  with  a-f  or 
video  amplifiers  is  negative  feedback  pro- 
portional to  output  voltage.  Figure  28  gives 
the  general  operating  conditions  for  feedback 
amplifiers.  Note  that  the  reduction  in  distor- 
tion is  proportional  to  the  reduction  in  gain  of 
the  amplifier,  also  that  the  reduction  in  the 
output  impedance  of  the  amplifier  is  somewhat 
greater  than  the  reduction  in  the  gain  by  an 
amount  which  is  a function  of  the  ratio  of  the 


H.  W.  Bods,  "Network  Analysis  and  Feedback  Ampli 
fier  Design,**  D.  Van  Nostrand  Co.,  250  Fourth  Ave, 
New  York  3,  N.  Y- 


130  Vacuum  Tube  Amplifiers 


THE  RADIO 


V,  Ra  Va 


DB  FEEDBACK  = 20  LOS  L Rg  + fiA^MVa  Roj 


; 20  LOS  ^ +R*[^'g^TAGE&Am_0F.J^2lj 
MN  OF  BOTH  STAGES  = f Gm^i  ( ^ p-* — ■)'|  * (&MVa  Rq) 

I,  Rb+Ha  J 


Rb  = 


X Rs 
Ri  + Ra 

R2 


OUTPUT  IMPEDANCE 


Ro 

RO  = REFLECTED  LOAD  IMPEDANCE  ON  Va 

Ra°  FEEDBACK  RESISTOR  (USUALLY  ABOUT  500  k) 

Rn  Ra 


[Rj  + Ra(6uv'2Ro)]x  0+  -g^) 


Rn  = PLATE  iMPEOANCe  OF  Va 


Figure  29  illustrates  a very  simple  and  ef- 
fective application  of  negative  voltage  feed- 
back to  an  output  pentode  or  tetrode  amplifier 
stage.  The  reduction  in  hum  and  distortion 
may  amount  to  15  to  20  db.  The  reduction  in 
the  effective  plate  impedance  of  the  stage  will 
be  by  a factor  of  20  to  100  dependent  upon  the 
operating  conditions.  The  circuit  is  commonly 
used  in  commercial  equipment  with  tubes  such 
as  the  6SJ7  for  Vj  and  the  6V6  or  6L6  for  V2. 

6-16  Vacuum-Tube  Voltmeters 

The  vacuum-tube  voltmeter  may  be  considered 
to  be  a vacuum-tube  detector  in  which  the 
rectified  d-c  current  is  used  as  an  indication 
of  the  magnitude  of  the  applied  alternating 
voltage.  The  vacuum  tube  voltmeter  (v.t.v.m.) 
consumes  little  or  no  power  and  it  may  be 
calibrated  at  60  cycles  and  used  at  audio  or 
radio  frequencies  with  little  change  in  the 
calibration. 


Figure  29 

SHUNT  FEEDBACK  CIRCUIT 
FOR  PENTODES  OR  TETRODES 
This  circuit  requires  only  the  addition  of 
one  resistor,  Rj,  to  the  normal  circuit  for 
such  an  application.  The  plate  impedance 
and  distortion  introduced  by  the  output 
stage  are  materially  reduced. 


output  impedance  of  the  amplifier  without 
feedback  to  the  load  impedance.  The  reduction 
in  noise  and  hum  in  those  stages  included 
within  the  feedback  loop  is  proportional  to  the 
reduction  in  gain.  However,  due  to  the  reduc- 
tion in  gain  of  the  output  section  of  the  ampli- 
fier somewhat  increased  gain  is  required  of 
the  stages  preceding  the  stages  included  with- 
in the  feedback  loop.  Therefore  the  noise  and 
hum  output  of  the  entire  amplifier  may  or  may 
not  be  reduced  dependent  upon  the  relative 
contributions  of  the  first  part  and  the  latter 
part  of  the  amplifier  to  hum  and  noise.  If  most 
of  the  noise  and  hum  is  coming  from  the  stages 
included  within  the  feedback  loop  the  un- 
desired signals  will  be  reduced  in  the  output 
from  the  complete  amplifier.  It  is  most  fre- 
quently true  in  conventional  amplifiers  that 
the  hum  and  distortion  come  from  the  latter 
stages,  hence  these  will  be  reduced  by  feed- 
back, but  thermal  agitation  and  microphonic 
noise  come  from  the  first  stage  and  will  not 
be  reduced  but  may  be  increased  by  feedback 
unless  the  feedback  loop  includes  the  first 
stage  of  the  amplifier. 


Basic  D-C  Vacuum-  A simple  v.t.v.m.  is 

Tube  Voltmeter  shown  in  figure  30. 

The  plate  load  may  be 
a mechanical  device,  such  as  a relay  or  a 
meter,  or  the  output  voltage  may  be  developed 
across  a resistor  and  used  for  various  con- 
trol purposes.  The  tube  is  biased  by  Ec  and 
a fixed  value  of  plate  current  flows,  causing 
a fixed  voltage  drop  across  the  plate  load 
resistor,  Rp.  When  a positive  d-c  voltage  is 
applied  to  the  input  terminals  it  cancels  part 
of  the  negative  grid  bias,  making  the  grid 
more  positive  with  respect  to  the  cathode. 
This  grid  voltage  change  permits  a greater 
amount  of  plate  current  to  flow,  and  develops 
a greater  voltage  drop  across  the  plate  load 
resistor.  A negative  input  voltage  would  de- 
crease the  plate  current  and  decrease  the 
voltage  drop  across  Rp.  The  varying  voltage 
drop  across  Rp  may  be  employed  as  a control 
voltage  for  relays  or  other  devices.  When  it  is 
desired  to  measure  various  voltages,  a voltage 


Figure  30 

SIMPLE  VACUUM  TUBE 
VOLTMETER 


HANDBOOK 


Vacuum  Tube  Voltmeters  131 


range  switch  (figure  31)  may  precede  the  v.t. 
v.m.  The  voltage  to  be  measured  is  applied  to 
voltage  divider,  Rl,  R2,  R3,  by  means  of  the 
"voltage  range’’  switch.  Resistor  R4  is  used 
to  protect  the  meter  from  excessive  input 
voltage  to  the  v.t.v.m.  In  the  plate  circuit  of 
the  tube  an  additional  battery  and  a variable 
resistor  ("zero  adjustment’’)  are  used  to 
balance  out  the  meter  reading  of  the  normal 
plate  current  of  the  tube.  The  zero  adjustment 
potentiometer  can  be  so  adjusted  that  the 
meter  M reads  zero  current  with  no  input  volt- 
age to  the  v.t.v.m.  When  a d-c  input  voltage 
is  applied  to  the  circuit,  cutrent  flows  through 
the  meter,  and  the  meter  reading  is  proportion- 
al to  the  applied  d-c  voltage. 

The  Bridge-type  Another  important  use 

V.T.V.M.  of  a d-c  amplifier  is  to 

show  the  exact  point  of 
balance  between  two  d-c  voltages.  This  is 
done  by  means  of  a bridge  circuit  with  two 
d-c  amplifiers  serving  as  two  legs  of  the 
bridge  (figure  32).  With  no  input  signal,  and 
with  matched  triodes,  no  current  will  be  read 
on  meter  M,  since  the  IR  drops  across  Ri  and 
R2  are  identical.  When  a signal  is  applied  to 
one  tube,  the  IR  drops  in  the  plate  circuits 
become  unbalanced,  and  meter  M indicates 
the  unbalance.  In  the  same  way,  two  d-c  volt- 
ages may  be  compared  if  they  are  applied  to 
the  two  input  circuits.  When  the  voltages  are 
equal,  the  bridge  is  balanced  and  no  current 
flows  through  the  meter.  If  one  voltage  changes, 
the  bridge  becomes  unbalanced  and  indication 
of  this  will  be  noted  by  a reading  of  the  meter. 

A Modern  VTVM  For  the  purpose  of  analysis, 
the  operation  of  a modern 
v.t.v.m.  will  be  described.  The  Heathkit  V-lA. 
is  a fit  instrument  for  such  a description,  since 
it  is  able  to  measure  positive  or  negative  d-c 
potentials,  a-c  r-m-s  values,  peak-to-peak 
values,  and  resistance.  The  circuit  of  this 
unit  is  shown  in  figure  33.  A sensitive  200  d-c 


Figure  32 

BRIDGE-TYPE  VACUUM  TUBE 
VOLTMETER 


microammeter  is  placed  in  the  cathode  circuit 
of  a 12AU7  twin  triode.  The  zero  adjust  control 
sets  up  a balance  between  the  two  sections  of 
the  triode  such  that  with  zero  input  voltage 
applied  to  the  first  grid,  the  voltage  drop  across 
each  portion  of  the  zero  adjust  control  is  the 
same.  Under  this  condition  of  balance  the 
meter  will  read  zero.  When  a voltage  is  applied 
to  the  first  grid,  the  balance  in  the  cathode 
circuits  is  upset  and  the  meter  indicates  the 
degree  of  unbalance.  The  relationship  between 
the  applied  voltage  on  the  first  grid  and  the 
meter  current  is  linear  and  therefore  the  meter 
can  be  calibrated  with  a linear  scale.  Since 
the  tube  is  limited  in  the  amount  of  current 
it  can  draw,  the  meter  movement  is  elec- 
tronically protected. 

The  maximum  test  voltage  applied  to  the 
12AU7  tube  is  about  3 volts.  Higher  applied 
voltages  are  reduced  by  a voltage  divider 
which  has  a total  resistance  of  about  10 
megohms.  An  additional  resistance  of  1-megohm 
is  located  in  the  d-c  test  prod,  thereby  per- 
mitting measurements  to  be  made  in  high  im- 
pedance circuits  with  minimum  disturbance. 

The  rectifier  portion  of  the  v.t.v.m.  is  shown 
in  figure  34.  When  a-c  measurements  are  de- 
sired, a 6AL5  double  diode  is  used  as  a full 
wave  rectifier  to  provide  a d-c  voltage  pro- 
pottionalto  the  applied  a-c  voltage.  This  d-c 
voltage  is  applied  through  the  voltage  divider 
string  to  the  12AU7  tube  causing  the  meter  to 
indicate  in  the  manner  previously  described. 
The  a-c  voltage  scales  of  the  meter  are  cali- 
brated in  both  RMS  and  peak-to-peak  values. 
In  the  1.5,  5,  15,  50,  and  150  volt  positions 
of  the  range  switch,  the  full  a-c  voltage  being 
measured  is  applied  to  the  input  of  the  6AL5 
full  wave  rectifier.  On  the  500  and  1500  volt 
positions  of  the  range  switch,  a divider  net- 
work reduces  the  applied  voltage  in  order  to 
limit  the  voltage  input  to  the  6AL5  to  a safe 
recommended  level. 


132 


THE  RADIO 


HANDBOOK 


Vacuum  Tube  Voltmeters  133 


TO  VTVM 
DC  I NPUT 
JACK 


Figure  34 

FULL-WAVE  RECTIFIER 
FOR  V.T.V.M. 


The  d’C  calibrate  control  (figure  33)  is  used 
to  obtain  the  proper  meter  deflection  for  the 
applied  a-c  voltage.  Vacuum  tubes  develop  a 
contact  potential  between  tube  elements.  Such 
contact  potential  developed  in  the  diode  would 
cause  a slight  voltage  to  be  present  at  all 
times.  This  voltage  is  cancelled  out  by  proper 
application  of  a bucking  voltage.  The  amount 
of  bucking  voltage  is  controlled  by  the  a-c 
balance  control.  This  eliminates  zero  shift 
of  the  meter  when  switching  from  a-c  to  d-c 
readings. 

For  resistance  measurements,  a 1.5  volt 
battery  is  connected  through  a string  of  multi- 
pliers and  the  external  resistance  to  be  meas- 
ured, thus  forming  a voltage  divider  across 
the  battery,  and  a resultant  portion  of  the 
battery  voltage  is  applied  to  the  12AU7  twin 


5 H I ELDED  PROBE  CAS E 


COAXIAL  LINE 

4,7  MEG 

T 

CK-705  m 

+ 

Figure  35 

R-F  PROBE  SUITABLE 
FOR  USE  IN  IKC-100  MC 
RANGE 


triode.  The  meter  scale  is  calibrated  in  re- 
sistance (ohms)  for  this  function. 

Test  Probes  Auxiliary  test  probes  may 

be  used  with  the  v.t.v.m. 
to  extend  the  operating  range,  or  to  measure 
radio  frequencies  with  high  accuracy.  Shown 
in  figure  35  is  a radio  frequency  probe  which 
provides  linear  response  to  over  100  mega- 
cycles. A crystal  diode  is  used  as  a rectifier, 
and  d-c  isolation  is  provided  by  a .005  uufd 
capacitor.  The  components  of  the  detector  are 
mounted  within  a shield  at  the  end  of  a length 
of  coaxial  line,  which  terminates  in  the  d-c 
input  jack  of  the  v.t.v.m.  The  readings  ob- 
tained are  RMS,  and  should  be  multiplied  by 
1.414  to  convert  to  peak  readings. 


CHAPTER  SEVEN 


High  Fidelity  Techniques 


The  art  and  science  of  the  reproduction  of 
sound  has  steadily  advanced,  following  the 
major  audio  developments  of  the  last  decade. 
Public  acceptance  of  home  music  reproduction 
on  a "high  fidelity”  basis  probably  dates  from 
the  summer  of  1948  when  the  Columbia  L-P 
microgroove  recording  techniques  were  intro- 
duced. 

The  term  high  fidelity  refers  to  the  repro- 
duction of  sound  in  which  the  different  dis- 
tortions of  the  electronic  system  are  held  below 
limits  which  are  audible  to  the  majority  of 
listeners.  The  actual  determination,  therefore, 
of  the  degree  of  fidelity  of  a music  system  is 
largely  psychological  as  it  is  dependent  upon 
the  ear  and  temperament  of  the  listener.  By  and 
large,  a rough  area  of  agreement  exists  as  to 
what  boundaries  establish  a "hi-fi”  system.  To 
enumerate  these  boundaries  it  is  first  necessary 
to  examine  sound  itself. 

7-1  The  Nature  of  Sound 

Experiments  with  a simple  tuning  fork  in 
the  seventeenth  century  led  to  the  discovery 
that  sound  consists  of  a series  of  condensations 
and  rarefactions  of  the  air  brought  about  by 
movement  of  ait  molecules.  The  vibrations  of 
the  prongs  of  the  fork  are  communicated  to 
the  surrounding  air,  which  in  turn  transmits 
the  agitation  to  the  ear  drums,  with  the  result 
that  we  hear  a sound.  The  vibrating  fork  pro- 
duces a sound  of  extreme  regularity,  and  this 
regularity  is  the  essence  of  music,  as  opposed  to 
noise  which  has  no  such  regularity. 


As  shown  in  figure  1,  the  sound  wave  of 
the  fork  has  frequency,  period,  and  pitch.  The 
frequency  is  a measure  of  the  number  of  vi- 
brations per  second  of  the  sound.  A fork  tuned 
to  produce  261  vibrations  per  second  is  tuned 
to  the  musical  note  of  middle-C.  It  is  of  in- 
terest to  note  that  any  object  vibrating,  moving, 
or  alternating  261  times  per  second  will  pro- 
duce a sound  having  the  pitch  of  middle-C. 
The  pitch  of  a sound  is  that  property  which  is 
determined  by  the  frequency  of  vibration  of 
the  source,  and  not  by  the  source  itself.  Thus 
an  electric  dynamo  producing  261  c.p.s.  will 
have  a hum-pitch  of  middle-C,  as  will  a siren, 
a gasoline  engine,  or  other  object  having  the 
same  period  of  oscillation. 


Figure  1 

VIBRATION  OF  TUNING  FORK  PRO- 
DUCES A SERIES  OF  CONDENSATIONS 
AND  RAREFACTIONS  OF  AIR  MOLE- 
CULES. THE  DISPLACEMENT  OF  AIR 
MOLECULES  CHANGES  CONTINUALLY 
WITH  RESPECT  TO  TIME,  CREATING 
A SINE  WAVE  OF  MOTION  OF  THE 
DENSITY  VARIATIONS. 


134 


Nature  of  Sound  135 


FREQUENCY  (cycles  per  seconb) 

NOTE 

c 

D 

E 

F 

G 

A 

B 

C' 

EQUAL- 

TEMPERED 

SCALE 

261.6 

293.7 

329.6 

349.2 

392.0 

440.0 

493.9 

523.2 

Figure  2 

THE  EQUAL-TEMPERED  SCALE  CON- 
TAINS TWELVE  INTERVALS,  EACH  OF 
WHICH  IS  1.06  TIMES  THE  FREQUEN- 
CY OF  THE  NEXT  LOWEST.  THE  HALF- 
TONE INTERVALS  INCLUDE  THE 
ABOVE  NOTES  PLUS  FIVE  ADDITION- 
AL NOTES:  277.2,  311.1,  370,  415.3, 
466.2  REPRESENTED  BY  THE  BLACK 
KEYS  OF  THE  PIANO. 


The  Musical  The  musical  scale  is  composed 
Seale  of  notes  or  sounds  of  various 

frequencies  that  bear  a pleasing 
aural  relationship  to  one  another.  Certain  com- 
binations of  notes  are  harmonious  to  the  ear 
if  their  frequencies  can  be  expressed  by  the 
simple  ratios  of  1:2,  2:3,  3:4,  and  4:5.  Notes 
differing  by  a ratio  of  1:2  are  said  to  be  sep- 
arated by  an  octave. 

The  frequency  interval  represented  by  an 
octave  is  divided  into  smaller  intervals,  form- 
ing the  musical  scales.  Many  types  of  scales 
have  been  proposed  and  used,  but  the  scale  of 
the  piano  has  dominated  western  music  for  the 
last  hundred  or  so  years.  Adapted  by  J.  S. 
Bach,  the  equal-tempered  scale  (figure  2)  has 
twelve  notes,  each  differing  from  the  next  by 
the  ratio  1:1.06.  The  reference  frequency,  or 
American  Standard  Pitch  is  A,  or  440-0  cycles. 

Harmonics  and  The  complex  sounds  pro- 
Overtones  duced  by  a violin  or  a wind 

instrument  beat  little  resem- 
blance to  the  simple  sound  wave  of  the  tuning 
fork.  A note  of  a clarinet,  for  example  (when 
viewed  on  an  oscilloscope)  resembles  figure  3. 
Vocal  sounds  are  even  mote  complex  than  this. 
In  1805  Joseph  Fourier  advanced  his  monu- 
mental theorem  that  made  possible  a mathe- 
matical analysis  of  all  musical  sounds  by  show- 
ing that  even  the  most  complex  sounds  are 
made  up  of  fundamental  vibrations  plus  har- 
monics, or  overtones.  The  tonal  qualities  of 
any  musical  note  may  be  expressed  in  terms  of 
the  amplitude  and  phase  relationship  between 
the  overtones  of  the  note. 

To  produce  overtones,  the  sound  source  must 
be  vibrating  in  a complex  manner,  such  as  is 
shown  in  figure  3.  The  resulting  vibration  is 

a combination  of  simple  vibrations,  producing 

a rich  tone  having  fundamental,  the  octave 
tone,  and  the  higher  overtones.  Any  sound  — 


INSTRUMENT  IS  A COMBINATION  OF 
SIMPLE  SINE-WAVE  SOUNDS,  CALLED 
HARMONICS.  THE  SOUND  OF  LOWEST 
FREQUENCY  IS  TERMED  THE  FUNDA- 
MENTAL. THE  COMPLEX  VIBRATION 
OF  A CLARINET  REED  PRODUCES  A 
SOUND  SUCH  AS  SHOWN  ABOVE. 


no  matter  how  complex  — can  be  analyzed 
into  pure  tones,  and  can  be  reproduced  by  a 
group  of  sources  of  pure  tones.  The  number 
and  degree  of  the  various  harmonics  of  a tone 
and  their  phase  relationship  determine  the 
quality  of  the  tone. 

For  reproduction  of  the  highest  quality,  these 
overtones  must  be  faithfully  reproduced.  A mu- 
sical note  of  523  cycles  may  be  rich  in  twen- 
tieth order  overtones.  To  reproduce  the  origi- 
nal quality  of  the  note,  the  audio  system  must 
be  capable  of  passing  overtone  frequencies  of 
the  order  of  11,000  cycles.  Notes  of  higher 
fundamental  frequency  demand  that  the  audio 
system  be  capable  of  good  reproduction  up  to 
the  maximum  response  limit  of  the  human 
ear,  in  the  region  of  15,000  cycles. 

Reproduction  Many  factors  enter  into  the 
Limitations  problem  of  high  quality  audio 
reproduction.  Most  important 
of  these  factors  influence  the  overall  design  of 
the  music  system.  These  are: 

1 —  Restricted  frequency  range. 

2 —  Nonlinear  distortions. 

3 —  Transient  distortion. 

4 —  Nonlinear  frequency  response. 

5 —  Phase  distortion. 

6 —  Noise,  "wow”,  and  "flutter”. 

A restricted  frequency  range  of  reproduction 
will  tend  to  make  the  music  sound  "tinny” 
and  unrealistic.  The  fundamental  frequency 
range  covered  by  the  various  musical  instru- 
ments and  the  human  voice  lies  between  15 
cycles  and  9,000  cycles.  Overtones  of  the  in- 
struments and  the  voice  extend  the  upper 
audible  limit  of  the  music  range  to  15,000 
cycles  or  so.  In  order  to  fully  reproduce  the 
musical  tones  falling  within  this  range  of  fre- 
quencies the  music  system  must  be  capable  of 
flawlessly  reproducing  all  frequencies  within 
the  range  without  discrimination. 


136  High  Fidelity  Techniques 


THE  RADIO 


BASIC  LIMITS  FOR  HIGH  FIDELITY  AND 
"GOOD  QUALITY''  REPRODUCTION 

TYPE  OF 

DISTORTION 

LIMIT 

HIGH  FIDELITY 

"GOOD'  REPRODUCTION 

RESTR ICTED 
FREQUENCY 

RANGE 

20-  15000  CPS 

50-  10000  CPS 

INTERMODULATION 
DISTORTION  AT 

FULL  OUTPUT 

4 

10  Vo 

HARMONIC 
DISTORTION  AT 

FULL  OUTPUT 

2 

5 Vo 

Y WOW" 

0.1  Vo 

t Vo 

HUM  AND  NOISE 

-70  OB  BELOW 
FULL  OUTPUT 

-50  08  BELOW 
FULL  OUTPUT 

Figure  4 


Nonlinear  qualities  such  as  harmonic  and 
intermodulation  (IM)  distortion  are  extremely 
objectionable  and  are  created  when  the  output 
of  the  music  system  is  not  exactly  proportional 
to  the  input  signal.  Nonlinearity  of  any  part 
of  the  system  produces  spurious  harmonic  fre- 
quencies, which  in  turn  lead  to  unwanted  beats 
and  resonances.  The  combination  of  harmonic 
frequencies  and  intermodulation  products  pro- 
duce discordant  tones  which  are  disagreeable 
to  the  ears. 

The  degree  of  intermodulation  may  be  meas- 
ured by  applying  two  tones  h and  of  known 
amplitude  to  the  input  of  the  amplifier  under 
test.  The  relative  amplitude  of  the  difference 
tone  (f:-fi)  is  considered  a measure  of  the 
intermodulation  distortion.  Values  of  the  order 
of  A%  IM  or  less  define  a high  fidelity  music 
amplifier. 

Response  of  the  music  system  to  rapid  tran- 
sient changes  is  extremely  important.  Tran- 
sient peaks  cause  overloading  and  shock-excita- 
tion of  resonant  circuits,  leaving  a "hang-over” 
effect  that  masks  the  clarity  of  the  sound.  A 
system  having  poor  transient  response  will  not 
sound  natural  to  the  ear,  even  though  the  dis- 
tortion factors  are  acceptably  low. 

Linear  frequency  response  and  good  power 
handling  capability  over  the  complete  audio 
range  go  hand  in  hand.  The  response  should 
be  smooth,  with  no  humps  or  dips  in  the  curve 
over  the  entire  frequency  range.  This  require- 
ment is  particularly  important  in  the  electro- 
mechanical components  of  the  music  system, 
such  as  the  phonograph  pickup  and  the  loud- 
speaker. 

Phase  distortion  is  the  change  of  phase  an- 
gle between  the  fundamental  and  harmonic 
frequencies  of  a complex  tone.  The  output 


wave  envelope  therefore  is  different  from  the 
envelope  of  the  input  wave.  In  general,  phase 
distortion  is  difficult  to  hear  in  sounds  having 
complex  waveforms  and  may  be  considered  to 
be  sufficiently  low  in  value  if  the  IM  figure  of 
the  amplifier  is  acceptable. 

Noise  and  distortions  introduced  into  the 
program  material  by  the  music  system  must 
be  kept  to  a minimum  as  they  are  particularly 
noticeable.  Record  scratch,  turntable  "rumble”, 
and  "flutter”  can  mar  an  otherwise  high  qual- 
ity system.  Inexpensive  phonograph  motors  do 
not  run  at  constant  speed,  and  the  slight  varia- 
tions in  speed  impart  a variation  in  pitch 
{wow)  to  the  music  which  can  easily  be  heard. 
Vibration  of  the  motor  may  be  detected  by  the 
pickup  arm,  superimposing  a low  frequency 
rumble  on  the  music. 

The  various  distortions  that  appear  in  a 
music  system  are  summarized  in  figure  4,  to- 
gether with  suggested  limits  within  which  the 
system  may  truly  be  termed  "high  fidelity.” 

7-2  The  Phonograph 

The  modern  phonograph  record  is  a thin 
disc  made  of  vinylite  or  shellac  material.  Disc 
rotation  speeds  of  78.3,  33  1/3,  and  45  r.p.m. 
are  in  use,  with  the  older  78.3  r.p.m.-  speed 
gradually  being  replaced  by  the  lower  speeds. 
A speed  of  16  2/3  r.p.m.  is  used  for  special 
"talking  book”  recordings.  A continuous  groove 
is  cut  in  the  record  by  the  stylus  of  the  record- 
ing machine,  spiralling  inward  towards  the 
center  of  the  record.  Amplitude  variations  in 
this  groove  proportional  to  the  sound  being 
recorded  constitute  the  means  of  placing  the 
intelligence  upon  the  surface  of  the  record. 
The  old  78.3  r.p.m.  recordings  were  cut  ap- 
proximately 100  grooves  per  inch,  while  the 
newer  "micro-groove”  recordings  are  cut  ap- 
proximately 250  grooves  per  inch.  Care  must 
be  taken  to  see  that  the  amplitude  excursions 
of  one  groove  do  not  fall  into  the  adjoining 
groove.  The  groove  excursions  may  be  con- 
trolled by  the  system  of  recording,  and  by 
equalization  of  the  recording  equipment. 

Recording  The  early  commercial  phono- 
Techniques  graph  records  were  cut  with  a 
mechanical-acoustic  system  that 
produced  a constant  velocity  characteristic  with 
the  amplitude  of  cut  increasing  as  the  recorded 
frequency  decreased  (figure  5A).  When  the 
recording  technique  became  advanced  enough 
to  reproduce  low  audio  frequencies,  it  was 
necessary  to  reduce  the  amplitude  of  the  lower 
frequencies  to  prevent  overcutting  the  record. 
A crossover  point  near  500  cycles  was  chosen. 


HANDBOOK 


High  Fidelity  Amplifier  137 


FREQUENCY  (CPs) 


CONSTANT  VELOCITY  RECOROtNO 


@ 


CONSTANT  AMPLITUDE  BELOW  CROSSOVER 
FREQUENCY,  HIGH  FREQUENCY  PRE-EMPHASiS 
ABOVE  CROSSOVER  FREQUENCY 


© 


and  a constant  amplitude  groove  was  cut  below 
this  frequency  (figure  5B).  This  system  does 
not  reproduce  rhe  higher  audio  notes,  since  the 
recording  level  rapidly  drops  into  the  surface 
noise  level  of  the  record  as  the  cutting  fre- 
quency is  raised.  The  modern  record  employs 
pre-emphasis  of  the  higher  frequencies  to  boost 
them  out  of  the  noise  level  of  the  record 
(figure  5C).  When  such  a record  is  played 
back  on  properly  compensated  equipment,  the 

audio  level  will  remain  well  above  the  back- 
ground noise  level,  as  shown  in  figure  5D, 


FREQUENCY  (cPS) 


CONSTANT  AMPLITUDE  BELOW  CROSSOVER 
FREQUENCY,  CONSTANT  VELOCITY  ABOVE 
CROSSOVER  FREQUENCY 


FREQUENCY  (cps) 


RESPONSE  OF  RECORD  OF  SC  PLAYED  ON 
PROPERLY  COMPENSATED  EQUIPMENT 


Figure  5 

MODERN  PHONOGRAPH  RECORD  EM- 
PLOYS CONSTANT  AMPLITUDE  CUT 
BELOW  CROSSOVER  POINT  AND  HIGH 
FREQUENCY  PRE-EMPHASIS  (BOOST) 
ABOVE  CROSSOVER  FREQUENCY. 


The  Phonograph  The  most  popular  types 

Pickup  of  pickup  cartridges  in 

use  today  are  the  high 
impedance  crystal  unit,  and  the  low  im- 
pedance variable  reluctance  cartridge.  The 
crystal  pickup  consists  of  a Rochelle  salt 
element  which  is  warped  by  the  action  of 
the  phonograph  needle,  producing  an  elec- 
trical impulse  whose  frequency  and  ampli- 
tude are  proportional  to  the  modulation  of 
the  record  groove.  One  of  the  new  "transducer” 
crystal  cartridges  is  shown  in  figure  6.  When 
working  into  a high  impedance  load,  the  out- 
put of  a high  quality  crystal  pickup  is  of  the 


138  High  Fidelity  Techniques 


THE  RADIO 


Figure  6 

NEW  CRYSTAL  "TRANSDUCER" 
CARTRIDGE  PROVIDES  HIGH- 
FIDELITY  OUTPUT  AT  RELATIVELY 
HIGH  LEVEL 


order  of  one-half  volt  or  so.  Inexpensive  crystal 
units  used  in  78  r.p.m.  record  changers  and 
ac-dc  phonographs  may  have  as  much  as  two 
or  three  volts  peak  output.  The  frequency  re- 
sponse of  a typical  high  quality  crystal  pickup 
is  shown  in  figure  7. 

The  variable  reluctance  pickup  is  shown  in 
figure  8.  The  reluctance  of  the  air  gap  in  a 
magnetic  circuit  is  changed  by  the  movement 
of  the  phonograph  needle,  creating  a variable 
voltage  in  a small  coil  coupled  to  the  magnetic 
lines  of  force  of  the  circuit.  The  output  im- 
pedance of  the  reluctance  cartridge  is  of  the 
order  of  a few  hundred  ohms,  and  the  output 
is  approximately  10  millivolts. 

For  optimum  performance,  an  equalized  pre- 
amplifier stage  is  usually  employed  with  the 
reluctance  pickup.  The  circuit  of  a suitable  unit 
is  shown  in  figure  9-  Equalization  is  provided 
by  Rs,  Ri,  and  Ca,  with  a low  frequency  cross- 
over at  about  500  cycles.  Total  equalization  is 
1 5 db.  High  frequency  response  may  be  limited 
by  reducing  the  value  of  Ri  to  5,000  — 15,000 
ohms. 

The  standard  pickup  stylus  for  78  r.p.m. 
records  has  a tip  radius  of  .0025  inch,  whereas 
the  microgroove  (33  1/3  and  45  r.p.m.)  stylus 
has  a tip  radius  of  .001  inch.  Many  pickups. 


FREOUENCY  (cps) 

Figure  7 

FREQUENCY  RESPONSE  OF  HIGH- 
QUALITY  CRYSTAL  PHONOGRAPH 
CARTRIDGE.  (ELECTROVOICE  56-DS 
POWER  POINT  TRANSDUCER) 


Figure  8 

"RELUCTANCE"  CARTRIDGE  IS 
STANDARD  PICK-UP  FOR 
MUSIC  SYSTEM. 

Low  stylui  pressure  oi  four  grams  insures 
minimum  record  wear.  Dual  stylus  is  used 
having  two  needle  tip  diameters  for  long 
playing  and  78  R.P.M.  recordings. 


therefore,  are  designed  to  have  interchangeable 
cartridges  or  needles  to  accomodate  the  dif- 
ferent groove  widths. 

7-3  The  High  Fidelity  Amplifier 

A block  diagram  of  a typical  high  fidelity 
system  is  shown  in  figure  10.  A preamplifier 
is  used  to  boost  the  output  level  of  the  phono- 
graph pickup,  and  to  permit  adjustment  of  in- 
put selection,  volume,  record  compensation, 
and  tone  control.  The  preamplifier  may  be 
mounted  directly  at  the  phonograph  turntable 
position,  permitting  the  larger  power  amplifier 
to  be  placed  in  an  out  of  the  way  position. 

The  power  amplifier  is  designed  to  operate 
from  an  input  signal  of  a volt  or  so  derived 
from  the  preamplifier,  and  to  build  this  signal 
to  the  desired  power  level  with  a minimum 
amount  of  distortion.  Maximum  power  output 
levels  of  ten  to  twenty  watts  are  common  for 
home  music  systems. 

The  power  supply  provides  the  smoothed, 
d-c  voltages  necessary  for  operation  of  the  pre- 
amplifier and  power  amplifier,  and  also  the 


6SC7 


Figure  9 

PREAMPLIFIER  SUITABLE  FOR  USE  WITH 
LOW  LEVEL  RELUCTANCE  CARTRIDGE. 


HANDBOOK 


High  Fidelity  Amplifier  139 


11  5 V. 


Figure  10 

BLOCK  DIAGRAM  OF  HIGH  FIDELITY 
MUSIC  SYSTEM. 


BASS  BOOST 
AND  ATTENUATION 


TREBLE  BOOST 
AND  ATTENUATION 


Figure  1 1 

SIMPLE  R-C  CIRCUITS  MAY  BE 
USED  FOR  BASS  AND  TREBLE 
BOOST  OR  ATTENUATION. 


filament  voltages  (usually  a-c)  for  the  heaters 
of  the  various  amplifier  tubes. 

The  loudspeaker  is  a device  which  couples 
the  electrical  energy  of  the  high  fidelity  sys- 
tem to  the  human  ear  and  usually  limits  the 
overall  fidelity  of  the  complete  system.  Great 
advances  in  speaker  design  have  been  made  in 
the  past  years,  permitting  the  loudspeaker  to 


Figure  12 

FREQUENCY  RESPONSE  CURVES  FOR 
THE  BASS  AND  TREBLE  BOOST 
AND  ATTENUATION  CIRCUITS 
OF  FIGURE  11. 


hold  its  own  in  the  race  for  true  fidelity. 
Speaker  efficiency  runs  from  about  10%  for 
cone  units  to  nearly  40%  for  high  frequency 
tweeters.  The  frequency  response  of  any  speak- 
er is  a function  of  the  design  and  construction 
of  the  speaker  enclosure  or  cabinet  that  mounts 
the  reproducer. 

Tone  Equalizer  networks  are  em- 

Compensot-on  ployed  in  high  fidelity  equip- 

ment to  1) — tailor  the  re- 
sponse curve  of  the  system  to  obtain  the  correct 
overall  frequency  response,  2 ) — to  compensate 
for  inherent  faults  in  the  program  material, 
3 ) — or  merely  to  satisfy  the  hearing  preference 
of  the  listener.  The  usual  compensation  net- 
works are  combinations  of  RC  and  RL  net- 
works that  provide  a gradual  attenuation  over 
a given  frequency  range.  The  basic  RC  net- 
works suitable  for  equlizer  service  are  shown 
in  figure  1 1 . Shunt  capacitance  is  employed 
for  high  frequency  attenuation,  and  series 
capacitance  is  used  for  low  frequency  atten- 
uation. A combination  of  these  simple  a-c 
voltage  dividers  may  be  used  to  provide  almost 
any  response,  as  shown  in  figure  12.  It  is 
common  practice  to  place  equalizers  between 
two  vacuum  tubes  in  the  low  level  stages  of 
the  preamplifier,  as  shown  in  figure  13.  Bass 
and  treble  boost  and  attenuation  of  the  order 


Figure  13 

BASS  AND  TREBLE  LEVEL 
CONTROLS,  AS 
EMPLOYED  IN  THE 
HEATH  KIT  WA-P2 
PREAMPLIFIER. 


BASS  CONTROL 


TREBLE  CONTROL 


SOUND  INTENSITY  LEVEL(db) 


140  High  Fidelity  Techniques 


THE  RADIO 


2NI90  2N190  -t2v.  2N190 


BASS  TREBLE 


Figure  14 

TRANSISTORIZED  HIGH-FIDELITY  PREAMPLIFIER  FOR  USE  WITH 
RELUCTANCE  PHONOGRAPH  CARTRIDGE. 


Figure  15 

THE  "FLETCHER-MUNSON"  CURVE 
ILLUSTRATING  THE  INTENSITY 
RESPONSE  OF  THE  HUMAN  EAR. 

of  15  db  may  be  obtained  from  such  a circuit. 
A simple  transistorized  preamplifier  using 
this  type  of  equalizing  network  is  shown  in 
figure  14. 


R1-R2-R31  THREE  SECTION  POTENTIOMETER,  IRC 
TYPE,  BUILT  OF  THE  FOLLOWINGi 
R 1 - IRC  PQ77-  733 
RZ-IRC  multisection  M73-737 
R3-//?C  multisection  M73-7Z8 

Figure  16 

VARIABLE  LOUDNESS  CONTROL  FOR 
USE  IN  LOW  IMPEDANCE  PLATE 
CIRCUITS.  MAY  BE  PURCHASED 
AS  IRC  TYPE  LC-1  LOUDNESS 
CONTROL. 


Loudness  The  minimum  threshold  of 

Compensolon  hearing  and  the  maximum 

threshold  of  pain  vary  great- 
ly with  the  frequency  of  the  sound  as  shown 
in  the  Fletcher-Munson  curves  of  figure  15. 
To  maintain  a reasonable  constant  tonal  bal- 
ance as  the  intensity  of  the  sound  is  changed 
it  is  necessary  to  employ  extra  bass  and  treble 
boost  as  the  program  level  is  decreased.  A 
simple  variable  loudness  control  is  shown  in 
figure  16  which  may  be  substituted  for  the 
ordinary  volume  control  used  in  most  audio 
equipment. 

The  Power  The  power  amplifier  stage  of 
Amplifier  the  music  system  must  supply 
driving  power  for  the  loud- 
speaker. Commercially  available  loudspeakers 
are  low  impedance  devices  which  present  a 


Figure  17 

IMPEDANCE  AND  FREQUENCY 
RESPONSE  OF  "4-OHM"  12-INCH 
SPEAKER  PROPERLY  MOUNTED  IN 
MATCHING  BAFFLE. 


varying  load  of  two  to  nearly  one  hundred  Shown  in  figure  18  is  a basic  push-pull 

ohms  to  the  output  stage  (figure  17).  It  is  triode  amplifier,  using  inverse  feedback  around 

necessary  to  employ  a high  quality  output  the  power  output  and  driver  stage.  A simple 

transformer  to  match  the  loudspeaker  load  to  triode  inverter  is  used  to  provide  180-degtee 

the  relatively  high  impedance  plate  circuit  of  phase  reversal  to  drive  the  grid  circuit  of  the 

the  power  amplifier  stage.  In  general,  push-  power  amplifier  stage.  Maximum  undistorted 

pull  amplifiers  are  employed  for  the  output  power  output  of  this  amplifier  is  about  8 watts, 

stage  since  they  have  even  harmonic  cancelling  A modification  of  the  basic  triode  amplifier 

properties  and  permit  better  low  frequency  is  the  popular  Williamson  circuit  (figure  19) 

response  of  the  output  transformer  since  there  developed  in  England  in  1947.  This  circuit 

is  no  d-c  core  saturation  effect  present.  rapidly  became  the  "standard  of  comparison” 

To  further  reduce  the  harmonic  distortion  in  a few  short  years.  Pentode  power  tubes  are 

and  intermodulation  inherent  in  the  amplifier  connected  as  triodes  for  the  output  stage,  and 

system  a negative  feedback  loop  is  placed  negative  feedback  is  taken  from  the  secondary 

around  one  or  more  stages  of  the  unit.  Fre-  of  the  output  transformer  to  the  cathode  of 

quency  response  is  thereby  improved,  and  the  the  input  stage.  Only  the  most  linear  portion 

output  impedance  of  the  amplifier  is  sharply  of  the  tube  characteristic  curve  is  used.  Al- 

reduced,  providing  a very  low  source  im-  though  that  portion  has  been  extended  by 

pedance  for  the  loudspeaker.  higher  than  normal  plate  supply  voltage,  it 


FEEDBACK  RESISTOR 


Figure  1 9 

U.  S.  VERSION  OF  BRITISH  "WILLIAMSON"  AMPLIFIER  PROVIDES  10  WATTS  POWER 
OUTPUT  AT  LESS  THAN  2%  INTERMODULATION  DISTORTION.  6SN7  STAGE  USES 

DIRECT  COUPLING. 


142  High  Fidelity  Techniques 


THE  RADIO 


FROM 

6SN7CT 

PHASE 

INVERTER 


807/5881 


TO  FEEDBACK 

CIRCUIT 


T 


Figure  20 


"ULTRA-LINEAR"  CONFIGURATION  OF  WILLIAMSON  AMPLIFIER  DOUBLES  POWER  OUT- 
PUT, AND  REDUCES  IM  LEVEL.  SCREEN  TAPS  ON  OUTPUT  TRANSFORMER  PERMIT 

"SEMI-TETRODE"  OPERATION. 


is  only  a fraction  of  the  curve  normally  used 
in  amplifiers.  Thus  a comparatively  low  output 
power  level  is  obtained  with  tubes  capable  of 
much  more  efficient  operation  under  less 
stringent  requirements.  With  400  volts  ap- 
plied to  the  output  stage,  a power  output  of  10 
watts  may  be  obained  wtih  less  than  2%  inter- 
modulation  distortion. 

A recent  variation  of  the  Williamson  cir- 
cuit involves  the  use  of  a tapped  output  trans- 
former. The  screen  grids  of  the  push-pull  am- 
plifier stage  are  connected  to  the  primary  taps, 
allowing  operating  efficiency  to  approach  that 
of  the  true  pentode.  Power  output  in  excess 
of  25  watts  at  less  than  2%  intermodulation  dis- 


Figure  21 

"BABY  HI-FI"  AMPLIFIER  IS  DWARFED 
BY  12-INCH  SPEAKER  ENCLOSURE 
This  miniature  music  system  is  capable  of  ex- 
cellent performance  in  the  small  home  or 
apartment.  Preamplifier^  bass  and  treble  con- 
trols, and  volume  control  are  all  incorporated 
in  the  unit.  Amplifier  provides  4 watts  output 
at  4%  IM  distortion. 


tortion  may  be  obtained  with  this  circuit 
(figure  20). 

7-4  Amplifier  Consfruction 

Wiring  Assembly  and  layout  of  high 

Techniques  fidelity  audio  amplifiers  fol- 

lows the  general  technique  des- 
cribed for  other  forms  of  electronic  equipment. 
Extra  care,  however,  must  be  taken  to  insure 
that  the  hum  level  of  the  amplifier  is  extreme- 
ly low.  A good  hi-fi  system  has  excellent  re- 
sponse in  the  60  cycle  region,  and  even  a 
minute  quantity  of  induced  a-c  voltage  will  be 
disagreeably  audible  in  the  loudspeaker.  Spur- 
ious eddy  currents  produced  in  the  chassis  by 
the  power  transformer  are  usually  responsible 
for  input  stage  hum. 

To  insure  the  lowest  hum  level,  the  power 
transformer  should  be  of  the  "upright"  type 
instead  of  the  "half-shell”  type  which  can 
couple  minute  voltages  from  the  windings  to 
a steel  chassis.  In  addition,  part  of  the  windings 
of  the  half-shell  type  project  below  the  chassis 
where  they  are  exposed  to  the  input  wiring  of 
the  amplifier.  The  core  of  the  power  trans- 
former should  be  placed  at  right  angles  to  the 
core  of  a nearby  audio  transformer  to  reduce 
spurious  coupling  between  the  two  units  to  a 
minimum. 

It  is  common  practice  in  amplifier  design 
to  employ  a ground  bus  return  system  for  all 
audio  tubes.  All  grounds  are  returned  to  a 
single  heavy  bus  wire,  which  in  turn  is 
grounded  at  one  point  to  the  metal  chassis. 
This  ground  point  is  usually  at  the  input  jack 
of  the  amplifier.  When  this  system  is  used, 
a-c  chassis  currents  are  not  coupled  into  the 
amplifying  stages.  This  type  of  construction  is 
illustrated  in  the  amplifiers  described  later  in 
this  chapter. 


HANDBOOK 


Amplifier  Construction  143 


Figure  22 

SCHEMATIC,  "BABY  HI-FI"  AMPLIFIER 

Ti — 260-0-260  volts  at  90  ma.,  6.3  volts  at  4.0  CHi — l.S  henry  at  200  mo.  Chicago  Standard  C-2327. 

amp.,  upright  mounting.  Chicago-Standard  CjA-B-C — 30-20-10  pfd.  350  volt.  Mallory  Fp-330.7 

PC-a420. 

Ti — 10  K,  CT.  to  8,  16  ohms.  Peerless  (Altec)  NOTE — Feedback  loop  returns  to  8 ohm  tap  on  Ts 

S-S10F.  when  8 ohm  speaker  is  used. 


Care  should  be  taken  to  reduce  the  capaci- 
tance to  the  chassis  of  high  impedance  circuits, 
or  the  high  frequency  response  of  the  unit  will 
suffer.  Shielded  "bath-tub”  type  capacitors 
should  not  be  used  for  interstage  coupling  ca- 
pacitors. Tubular  paper  capacitors  are  satis- 
factory. These  should  be  spaced  well  away  from 
the  chassis. 

It  is  a poor  idea  to  employ  the  chassis  as  a 
common  filament  return,  especially  for  low 
level  audio  stages.  The  filament  center-tap  of 
the  power  transformer  should  be  grounded, 
and  twisted  filament  wires  run  to  each  tube 
socket.  High  impedance  audio  components  and 
wiring  should  be  kept  clear  of  the  filament 
lines,  which  may  even  be  shielded  in  the  vicin- 
ity of  the  input  stage.  In  some  instances,  the 
filament  center  tap  may  be  taken  from  the  atm 
of  a low  resistance,  wirewound  potentiometer 
placed  across  the  filament  pins  of  the  input 
tube  socket.  The  arm  of  this  potentiometer  is 
grounded,  and  the  setting  of  the  control  is  ad- 
justed for  minimum  speaker  hum. 

7-5  The  "Baby  Hi  Fi" 

A definite  need  exists  for  a compact,  high 
fidelity  audio  amplifier  suitable  for  use  in  the 
small  home  or  apartment.  Listening  tests  have 
shown  that  an  average  power  level  of  less  than 


one  watt  in  a high  efficiency  speaker  will  pro- 
vide a comfortable  listening  level  for  a small 
room,  and  levels  in  excess  of  two  or  three  watts 
are  uncomfortably  loud  to  the  ear.  The  "Baby 
Hi-Fi”  amplifier  has  been  designed  for  use 
in  the  small  home,  and  will  provide  excellent 
quality  at  a level  high  enough  to  rattle  the 
windows. 

Designed  around  the  new  Electro-Voice  min- 
iature ceramic  cartridge,  the  amplifier  will  pro- 
vide over  4 watts  power,  measured  at  the  sec- 
ondary of  the  output  transformer.  At  this  level, 
the  distortion  figure  is  below  1%,  and  the  IM 
figure  is  4%.  At  normal  listening  levels,  the  IM 
is  much  lower,  as  shown  in  figure  24. 

The  Amplifier  The  schematic  of  the  ampli- 

Circuit  fier  is  shown  in  figure  22. 

Bass  and  treble  boost  con- 
trols are  incorporated  in  the  circuit,  as  is  the 
volume  control.  A dual  purpose  12AU7  double 
triode  serves  as  a voltage  amplifier  with  cath- 
ode degeneration.  A simple  voltage  divider 
network  is  used  in  the  grid  circuit  to  prevent 
amplifier  overloading  when  the  ceramic  cart- 
ridge is  used.  The  required  input  signal  for 
maximum  output  is  of  the  order  of  0-3  volts. 
The  output  level  of  the  Electro-Voice  cartridge 
is  approximately  twice  this,  as  shown  in  figure 
7.  The  use  of  the  high-level  cartridge  elimin- 


144  High  Fidelity  Techniques 


THE  RADIO 


Figure  23 
UNDER-CHASSIS 
VIEW  OF 
"BABY  HI-FI" 

Low  level  audio  stages  are  at 
upper  left,  with  components 
mounted  between  socket  pins 
and  potentiometer  controls. 
6X5  socket  is  at  lower  cen- 
ter of  photo  with  filter 
choke  CHi  at  right.  Feed- 
back resistor  Ri  is  at  left 
of  rectifier  socket* 


ates  the  necessity  of  high  gain  amplifiers  re- 
quired when  low  level  magnetic  pickup  heads 
ate  used.  Problems  of  hum  and  distortion  in- 
troduced by  these  extra  stages  ate  thereby 
eliminated,  greatly  simplifying  the  amplifier. 
The  second  section  of  the  12AU7  is  used  for 
bass  and  treble  boost.  Simple  R-C  networks 
are  placed  in  the  grid  circuit  permitting  gain 
boost  of  over  12  db  at  the  extremities  of  the 
response  range  of  the  amplifier. 

A second  12AU7  is  employed  as  a direct 
coupled  "hot-cathode”  phase  inverter,  capaci- 
tively  coupled  to  two  6AQ5  pentode  connected 


Figure  24 

INTERMODULATION  CURVE  FOR 
“BABY  HI-FI"  AS  MEASURED  ON 
HEATHKIT  INTERMODULATION 
ANALYZER. 


output  tubes.  The  feedback  loop  is  run  from 
the  secondary  of  the  output  transformer  to  the 
cathode  of  the  input  section  of  the  phase  in- 
verter. 

The  power  supply  of  the  "Baby  Hi-Fi”  con- 
sists of  a 6X5 -GT  rectifier  and  a capacitor  in- 
put filter.  A second  R-C  filter  section  is  used 
to  smooth  the  d-c  voltage  applied  to  the 
12AU7  tubes.  A cathode-type  rectifier  is  used 
in  preference  to  the  usual  filament  type  to  pre- 
vent voltage  surges  during  the  warm-up  period 
of  the  other  cathode-type  tubes. 

Amplifier  The  complete  amplifier  is 

Construction  built  upon  a small  "amplifier 

foundation”  chassis  and  cover 
measuring  5"x7"x6"  (Bud  CA-1754).  Height 
of  the  amplifier  including  dust  cover  is  6". 
The  power  transformer  (Ti)  and  output  trans- 
former (Tz)  are  placed  in  the  rear  corners  of 
the  chassis,  with  the  *6X5 -GT  rectifier  socket 
placed  between  them.  The  small  filter  choke 
(CHi)  is  mounted  to  the  wall  of  the  chassis 
and  may  be  seen  in  the  under-chassis  photo- 
graph of  figure  23.  The  four  audio  tubes  are 
placed  in  a row  across  the  front  of  the  chassis. 
Viewed  from  the  front,  the  12AU7  tubes  are 
to  the  left,  and  the  6AQ5  tubes  are  to  the  right. 
The  three  section  filter  capacitor  (GA,  B,  C) 
is  a chassis  mounting  unit,  and  is  placed  be- 
tween the  rectifier  tube  and  the  four  audio 
tubes.  Since  the  chassis  is  painted,  it  is  im- 
portant that  good  grounding  points  be  made  at 
each  tube  socket.  The  paint  is  cleared  away 


HANDBOOK 


"Baby  Hi-Fi"  145 


Figure  25 
TYPICAL 

IHTERMODULATION 
TEST  OF  AUDIO 
AMPLIFIER. 

Audio  tones  of  two  fre- 
quencies are  applied  to  in- 
put of  amplifier  under  fest« 
and  amplitude  of  "sum''  or 
"difference"*  frequency  is 
measured,  providing  relative 
inter-modulation  figure. 


beneath  the  socket  bolt  heads,  and  lock  nuts 
are  used  beneath  the  socket  retaining  nuts  to 
insure  a good  ground  connection.  All  ground 
leads  of  the  first  12AU7  tube  are  returned  to 
the  socket,  whereas  all  grounds  for  the  rest  of 
the  circuit  are  returned  to  a ground  lug  of 
filter  capacitor  Ci. 

Since  the  input  level  to  the  amplifier  is  of 
the  order  of  one-half  volt,  the  problem  of 
chassis  ground  currents  and  hum  is  not  so 
prevalent,  as  is  the  case  with  a high  gain  input 
stage. 

Phonograph-type  coaxial  receptacles  are 
mounted  on  the  rear  apron  of  the  chassis,  serv- 
ing as  the  input  and  output  connections.  The 
four  panel  controls  (bass  boost,  treble  boost, 
volume,  and  a-c  on)  are  spaced  equidistant 
across  the  front  of  the  chassis. 

Amplifier  The  filament  wiring  should  be 
Wiring  done  first.  The  center-tap  of  the 

filament  winding  is  grounded  to 
a lug  of  the  6X5-GT  socket  ring,  and  the  6-3 
volt  leads  from  the  transformer  are  attached 
to  pins  2 and  7 of  the  same  socket.  A twisted 
pair  of  wires  run  from  the  rectifier  socket  to 
the  right-hand  6AQ5  socket  (figure  23).  The 
filament  leads  then  proceed  to  the  next  6AQ5 
socket  and  then  to  the  two  12AU7  sockets  in 
turn. 

The  12AU7  preamplifier  stage  is  wired  next, 
A two  terminal  phenolic  tie-point  strip  is 
mounted  to  the  rear  of  the  chassis,  holding  the 
12K  decoupling  resistor  and  the  positive  lead 


of  the  10  gfd.,  450-volt  filter  capacitor.  All 
B-plus  leads  are  run  to  this  point.  Most  of  the 
components  of  the  bass  and  treble  boost  system 
may  be  mounted  between  the  tube  socket  ter- 
minals and  the  terminals  of  the  two  potentio- 
meters. The  feedback  resistor  Ri  is  mounted 
between  the  terminal  of  the  coaxial  output 
connector  and  a phenolic  tie-point  strip  placed 
beneath  an  adjacent  socket  bolt. 

When  the  wiring  has  been  completed  and 
checked,  the  amplifier  should  be  turned  on, 
and  the  various  voltages  compared  with  the 
values  given  on  the  schematic.  It  is  important 
that  the  polarity  of  the  feedback  loop  is  correct. 
The  easiest  way  to  reverse  the  feedback  polarity 


Figure  26 

20-WATT  "WILLIAMSON-TYPE 
AMPLIFIER  PROVIDES  ULTIMATE  IN 
LISTENING  PLEASURE  FOR  THE 
"GOLDEN  EAR." 

Amplifier  chassis  (left)  employs  two  low  level 
stages  driving  push-pull  807  tubes  in  so-called 
"Ultra-linear"  circuit.  Power  supply  is  at  right. 


146  High  Fidelity  Techniques 


3-  VOLTAGE  MEASUREMENTS  MADE 
WITH  1000  OHMS/VOLT  VOLTMETER 


7-  PUNCHED  AND  DRILLED  CHASSIS. 
STANCOP  tVM6 


4-  JACKS  Jl  AND  Jz  ARE  INSULATED 

PROM  CHASSIS  Figure  27 

SCHEMATIC  OF  20-WATT  MUSIC  SYSTEM  AMPLIFIER. 


is  to  cross-connect  the  two  plate  leads  of  the 
6AQ5  tubes.  If  the  feedback  polarization  is 
incorrect,  the  amplifier  will  oscillate  at  a super- 
sonic frequency  and  the  reproduced  signal  will 
sound  fuzzy  to  the  ear.  The  correct  connection 
may  be  determined  with  the  aid  of  an  oscillo- 
scope, as  the  oscillation  will  be  easily  found. 
The  builder  might  experiment  with  different 
values  of  feedback  resistor  Ri,  especially  if  a 
speaker  of  different  impedance  is  employed. 
Increasing  the  value  of  Ri  will  decrease  the 
degree  of  feedback.  For  an  8-ohm  speaker,  Ri 
should  be  decreased  in  value  to  maintain  the 
same  amount  of  feedback. 

This  amplifier  was  used  in  conjunction  with 
a General  Electric  S-1201A  12-inch  speaker 
mounted  in  an  Electro-Voice  KD6  Aristocrat 
speaker  enclosure  which  was  constructed  from 
a kit.  The  reproduction  was  extremely  smooth, 
with  good  balance  of  bass  and  treble. 

7-6  A High  Quality 
25  Watt  Amplifier 

This  amplifier  is  recommended  for  the  music 
lover  who  desires  the  utmost  in  high  fidelity. 
Capable  of  delivering  25  watts  output  at  less 
than  1.5%  intermodulation  distortion,  the  am- 
plifier will  provide  flawless  reproduction  at 
higher  than  normal  listening  levels.  It  is  true 
that  other  designs  have  been  advocated  pro- 
viding greater  power  at  a lower  IM  distortion 
level.  Flowever,  the  improvement  in  reproduc- 


tion of  such  a system  is  not  noticeable  — even 
to  the  trained  ear  — over  the  results  obtained 
with  this  low  priced  amplifier.  A full  20  watts 
of  audio  power  are  obtained  over  the  frequency 
range  of  20  to  50,000  cycles  at  less  than  0.1% 
harmonic  distortion.  At  the  average  listening 
level  of  1 watt,  the  frequency  response  is  plus 
or  minus  1 decibel  at  100,000  cycles,  assuring 
true  high  fidelity  response  over  the  entire  aud- 
ible range.  Preamplification  and  tone  compen- 
sation are  not  included  in  this  unit,  as  these 
funaions  are  accomplished  in  the  auxiliary 
driving  amplifier.  If  a variable  reluctance  cart- 
ridge is  employed,  the  preamplifiers  of  figure 
9 or  figure  14  may  be  used  with  this  unit. 
The  Heathkit  WA-P2  preamplifier  is  also 
recommended  for  general  use  when  several  in- 
put circuits  are  to  be  employed. 

The  Amplifier  The  schematic  of  this  ampli- 
Circuit  fier  is  shown  in  figure  27. 

It  is  a "Williamson-type” 
amplifier  using  the  so-called  "Ultra-linear” 
screen-tap  circuit  on  the  output  stage.  A dual 
triode  6SN7-GT  serves  as  a voltage  amplifier 
and  direct  coupled  phase  inverter.  Maximum 
signal  input  to  this  stage  is  approximately  0.8 
volt,  r.m.s.  for  rated  output  of  the  amplifier. 
The  feedback  loop  from  the  speaker  voice  coil 
is  returned  to  the  cathode  circuit  of  the  input 
stage,  placing  all  amplifying  stages  within  the 
loop. 

A second  6SN7-GT  is  used  as  a push-pull 


Figure  28 

UNDER-CHASSIS  VIEW 
OF  20-WATT  AMPLIFIER. 

807  SOCKETS  ARE  AT 
CENTER,  WITH  CATHODE 
BALANCING 
POTENTIOMETER  AT 
LOWER  RIGHT. 

Note  ground  bus,  storting  at 
center,  top  filter  capacitor,  and 
running  in  loop  around  under- 
side of  chassis.  Bus  is  grounded 
to  chassis  at  input  plug  (upper 
left).  Shielded  leads  run  to  vol- 
ume control  (center). 


driver  stage  for  the  high  level  output  amplifier. 
A plate  potential  of  430  volts  is  applied  to  this 
stage  to  insure  ample  grid  drive  to  the  output 
stage.  Large  value  coupling  capacitors  are  used 
between  all  stages  to  prevent  signal  degenera- 
tion at  the  lower  audio  frequencies.  Two  807 
beam-power  tubes  are  employed  in  the  final 
amplifier  stage-  Cathode  bias  is  applied  to  these 
tubes,  and  the  current  of  the  tubes  may  be 
equalized  by  potentiometer  Rs  in  the  bias 
circuit. 

Degenerative  feedback  is  taken  from  the  sec- 
ondary of  the  output  transformer  and  applied 
to  the  input  of  the  amplifier.  The  amount  of 
feedback  is  controlled  by  the  value  of  feed- 
back resistor  R.i.  A small  capacitor  is  placed 
across  Rs  to  reduce  any  tendency  towards  high 
frequency  parasitic  oscillation  sometimes  en- 
countered when  long  leads  are  used  to  connect 
the  amplifier  to  the  loudspeaker. 

An  external  power  supply  is  used  with  the 
amplifier,  delivering  440  volts  at  a current  of 
175  milliamperes.  and  6.3  volts  at  5 amperes. 

Amplifier  The  amplifier  is  built  upon  a 

Construction  steel  chassis  measuring  9”x 
7"x2".  Punched  and  drilled 
chassis  for  both  the  amplifier  and  power  sup- 
ply may  be  obtained  as  standard  parts,  as  speci- 
fied in  figure  27,  eliminating  the  necessity  of 
considerable  metal  work.  Placement  of  the 
major  components  may  be  seen  in  the  top  and 
under-chassis  photographs.  The  two  cathode 
current  jacks,  Ji  and  are  mounted  in  over- 
size holes  and  are  insulated  from  the  chassis 
with  fibre  shoulder  washers  placed  on  the  jack 
stem,  and  flat  fibre  washers  beneath  the  retain- 
ing nut.  The  can-type  filter  capacitors  are  in- 


sulated from  the  chassis  by  fibre  mounting 
boards. 

A common  ground  bus  is  used  in  this  am- 
plifier, and  all  "grounds"  are  returned  to  the 
bus,  rather  than  to  the  chassis.  The  bus  is 
grounded  at  the  input  jack  of  the  amplifier 
and  runs  around  the  underside  of  the  chassis, 
ending  at  the  ground  terminals  of  the  high 
voltage  filter  capacitors.  Tie-point  terminal 
strips  are  used  to  support  the  smaller  com- 
ponents, and  direct  point-to-point  wiring  is 
used. 

Matched  pairs  of  resistors  should  be  used 
in  the  balanced  audio  circuits,  and  these  resis- 
tors are  marked  ( * ) on  the  schematic.  If  possi- 
ble, measure  a quantity  of  resistors  in  the  store 
with  an  ohmmeter,  and  pick  out  the  pairs  of 
resistors  that  are  most  evenly  matched.  The 
exact  resistance  value  is  not  critical  within  10% 
as  long  as  the  two  resistors  are  close  in  value. 
Care  must  be  taken  when  these  resistors  are 
soldered  in  the  circuit,  as  the  heat  of  the  sol- 
dering iron  may  cause  the  resistance  value  to 
vary.  It  is  wise  to  grasp  the  resistor  lead  with 
a long-nose  pliers,  holding  the  lead  between 
the  point  where  the  iron  is  applied  and  the 
body  of  the  resistor.  The  pliers  will  act  as  a 
"heat  sink”,  protecting  the  resistor  from  exces- 
sive heat. 

The  filament  leads  are  made  up  of  a twisted 
pair  of  wires  running  between  the  various 
sockets.  Keep  the  twisted  leads  clear  of  the  in- 
put jack  to  insure  minimum  hum  pickup.  The 
plate  leads  from  the  807  caps  pass  through 
%-inch  rubber  grommets  mounted  in  the  chas- 
sis to  phenolic  tie  points  mounted  beneath  the 
chassis.  The  plate  leads  of  output  transformer 
Ti  attach  to  these  tie  points.  Be  sure  you  ob- 


IFgWWIPii 


148  High  Fidelity  Techniques 


Figure  29 

intermodulation  curve 
OF  20-watt  amplifier. 


serve  the  color  code  for  these  leads,  as  given 
in  figure  27,  as  the  proper  polarization  of  the 
leads  is  required  for  proper  feedback  operation. 

When  the  amplifier  wiring  is  completed,  all 
connections  should  be  checked  for  accidental 
grounds  or  transpositions.  Make  sure  the  meter 
jacks  are  insulated  from  the  chassis. 

The  Power  A companion  power  supply  for 
Supply  the  high  fidelity  amplifier  is 

shown  in  figure  30.  A cathode- 
type  5V4-G  rectifier  is  used  to  limit  the  warm- 
up surge  voltages  usually  encountered  in  such 


a supply.  Filament  and  plate  voltages  for  the 
Heathkit  W A-P2  preamplifier  may  be  derived 
from  receptacle  SO-1.  If  the  preamplifier  is  not 
used,  or  if  another  type  of  preamplifier  is  to 
be  employed,  the  center-tap  of  the  6.3  volt 
filament  winding  of  power  transformer  Ti 
(green/yellow  lead)  should  be  grounded  to 
the  chassis  of  the  power  supply. 

Wiring  of  this  unit  is  straightforward,  and 
no  unusual  precautions  need  be  taken.  It  is 
wise  to  attach  the  amplifier  to  the  supply  be- 
fore it  is  mrned  to  limit  warm-up  voltage 
excursions. 

Amplifier  The  amplifier  is  attached  to  the 

Operation  power  supply  by  a short  length 

of  4-wire  cable.  Make  sure  taht 
the  filament  leads  (pins  1 and  4,  Pi)  are 
made  of  sufficiently  heavy  wire  to  insure  that 
the  filaments  of  the  amplifier  receive  a full 
6.3  volts.  The  amplifier  should  be  turned  on 
and  all  voltages  checked  against  figure  27.  A 
speaker  or  suitable  resistive  load  should  be  at- 
tached to  output  terminal  strip  Pi,  and  a 0-100 
d-c  milliammeter  plugged  into  jacks  Ji  and  Ji. 
Balance  potentiometer  Ri  is  now  adjusted  until 
the  two  readings  are  the  same.  Each  measure- 
ment will  be  very  close  to  60  milliamperes 
when  the  currents  are  balanced.  The  amplifier 
is  now  ready  for  operation,  and  has  a fidelity 
curve  similar  to  that  shown  in  figure  29- 


Tl  5V4-G 


NOTE:  IF  HEATHKIT  PREAMPLIFieR 

IS  NOr  USED,  CENTER-TAP  WIRE 
{GREEN-yELLOW)  OF  6.3  FILAMENT 
WINDING  OF  T1  MUST  BE  GROUNDED. 


SOCKET  FOR 
HEATH  WA-PZ 
PREAMPLIFIER 


Tl  - 400-0-400  VOLTS,  200  MA..  5 VOLTS. 
3 AMP.,  6.3  VOLTS.  SAMP. 
CHICAGO-STANDARD  PC-641Z 
CH  1 - 4.5  HENRY  AT  200  MA. 

CHICAGO-STANDARD  C-14II 


Figure  30 

POWER  SUPPLY  FOR  20-WATT  AMPLIFIER. 


CHAPTER  EIGHT 


Radio  Frequency 
Vacuum  Tube  Amplifiers 


TUNED  RF  VACUUM  TUBE  AMPLIFIERS 


Tuned  r-f  voltage  amplifiers  are  used  in  re- 
ceivers for  the  amplification  of  the  incoming 
r-f  signal  and  for  the  amplification  of  inter- 
mediate frequency  signals  after  the  incoming 
frequency  has  been  converted  to  the  intermed- 
iate frequency  by  the  mixer  stage.  Signal  fre- 
quency stages  are  normally  called  tuned  r-f 
amplifiers  and  intermediate-frequency  stages 
ate  called  i-f  amplifiers.  Both  tuned  r-f  and 
i-f  amplifiers  are  operated  Class  A and  nor- 
mally operate  at  signal  levels  from  a fraction 
of  a microvolt  to  amplitudes  as  high  as  10  to 
50  volts  at  the  plate  of  the  last  i-f  stage  in  a 
receiver. 

8-1  Grid  Circuit 

Considerations 

Since  the  full  amplification  of  a receiver  fol- 
lows the  first  tuned  circuit,  the  operating  con- 
ditions existing  in  that  circuit  and  in  its  cou- 
pling to  the  antenna  on  one  side  and  to  the 
grid  of  the  first  amplifier  stage  on  the  other 
are  of  greatest  importance  in  determining  the 
signal-to-noise  ratio  of  the  receiver  on  weak 
signals. 

First  Tuned  It  is  obvious  that  the  highest 
Circuit  ratio  of  signal-to-noise  be  im- 

pressed on  the  grid  of  the  first 
r-f  amplifier  tube.  Attaining  the  optimum  ratio 
is  a complex  problem  since  noise  will  be  gen- 
erated in  the  antenna  due  to  its  equivalent 
radiation  resistance  (this  noise  is  in  addition 
to  any  noise  of  atmospheric  origin)  and  in  the 


first  tuned  circuit  due  to  its  equivalent  cou- 
pled resistance  at  resonance.  The  noise  volt- 
age generated  due  to  antenna  radiation  resist- 
ance and  to  equivalent  tuned  circuit  resistance 
is  similar  to  that  generated  in  a resistor  due 
to  thermal  agitation  and  is  expressed  by  the 
following  equation: 

E„'  = 4/feTRAf 

Where;  £„  = r-m-s  value  of  noise  voltage  over 
the  interval  Af 

k — Boltzman’s  constant  = 1.374 
X 10’’^  joule  per  °K. 

T = Absolute  temperature  °K. 

R = Resistive  component  of  imped- 
ance across  which  thermal  noise 
is  developed. 

Af  = Frequency  band  across  which 
voltage  is  measured. 

In  the  above  equation  Af  is  essentially  the 
frequency  band  passed  by  the  intermediate  fre- 
quency amplifier  of  the  receiver  under  consid- 
eration. This  equation  can  be  greatly  simpli- 
fied for  the  conditions  normally  encountered 
in  communications  work.  If  we  assume  the  fol- 
lowing conditions:  T = 300°  K or  27°  C or 
80.5°  F,  room  temperature;  Af  = 8000  cycles 
(the  average  pass  band  of  a communications 
receiver  or  speech  amplifier),  the  equation  re- 
duces to:  Ej  s.  - 0.0115  microvolts.  Ac- 
cordingly, the  thermal-agitation  voltage  ap- 
pearing in  the  center  of  half-wave  antenna  (as- 
suming effective  temperature  to  be  300°  K) 
having  a radiation  resistance  of  73  ohms  is 


149 


150  R-F  Vacuum  Tube  Amplifiers 


THE  RADIO 


approximately  0.096  microvolts.  Also,  the  ther- 
mal agitation  voltage  appearing  across  a 500,- 
000-ohm  grid  resistor  in  the  first  stage  of  a 
speech  amplifier  is  approximately  8 microvolts 
under  the  conditions  cited  above.  Further,  the 
voltage  due  to  thermal  agitation  being  im- 
pressed on  the  grid  of  the  first  t-f  stage  in  a 
receiver  by  a first  tuned  circuit  whose  reson- 
ant resistance  is  50,000  ohms  is  approximately 
2.5  microvolts.  Suffice  to  say,  however,  that 
the  value  of  thermal  agitation  voltage  appear- 
ing across  the  first  tuned  circuit  when  the  an- 
tenna is  properly  coupled  to  this  circuit  will 
be  very  much  less  than  this  value. 

It  is  common  practice  to  match  the  imped- 
ance of  the  antenna  transmission  line  to  the 
input  impedance  of  the  grid  of  the  first  r-f  am- 
plifier stage  in  a receiver.  This  is  the  condi- 
tion of  antenna  coupling  which  gives  maximum 
gain  in  the  receiver.  However,  when  u-h-f  tubes 
such  as  acorns  and  miniatures  are  used  at  fre- 
quencies somewhat  less  than  their  maximum 
capabilities,  a significant  improvement  in  sig- 
nal-to-noise  ratio  can  be  attained  by  increas- 
ing the  coupling  between  the  antenna  and  first 
tuned  circuit  to  a value  greater  than  that  which 
gives  greatest  signal  amplitude  out  of  the  re- 
ceiver. In  other  words,  in  the  10,  6,  and  2 me- 
ter bands  it  is  possible  to  attain  somewhat  im- 
proved signal-to-noise  ratio  by  increasing  an- 
tenna coupling  to  the  point  where  the  gain  of 
the  receiver  is  slightly  reduced. 

It  is  always  possible,  in  addition,  to  obtain 
improved  signal-to-noise  ratio  in  a v-h-f  re- 
ceiver through  the  use  of  tubes  which  have 
improved  input  impedance  characteristics  at 
the  frequency  in  question  over  conventional 
types. 

Noise  Factor  The  limiting  condition  for  sen- 
sitivity in  any  receiver  is  the 
thermal  noise  generated  in  the  antenna  and  in 
the  first  tuned  circuit.  However,  with  proper 
coupling  between  the  antenna  and  the  grid  of 
the  tube,  through  the  first  tuned  circuit,  the 
noise  contribution  of  the  first  tuned  circuit 
can  be  made  quite  small.  Unfortunately,  though, 
the  major  noise  contribution  in  a properly  de- 
signed receiver  is  that  of  the  first  tube.  The 
noise  contribution  due  to  electron  flow  and 
due  to  losses  in  the  tube  can  be  lumped  into 
an  equivalent  value  of  resistance  which,  if 
placed  in  the  grid  circuit  of  a perfect  tube  hav- 
ing the  same  gain  but  no  noise  would  give  the 
same  noise  voltage  output  in  the  plate  load. 
The  equivalent  noise  resistance  of  tubes  such 
as  the  6SK7,  6SG7,  etc.,  tuns  from  5000  to 
10,000  ohms.  Very  high  Gn,  tubes  such  as  the 
6AC7  and  6AK5  have  equivalent  noise  resist- 
ances as  low  as  700  to  1500  ohms.  The  lower 
the  value  of  equivalent  noise  resistance,  the 


lower  will  be  the  noise  output  under  a fixed 
set  of  conditions. 

The  equivalent  noise  resistance  of  a tube 
must  not  be  confused  with  the  actual  input 
loading  resistance  of  a tube.  For  highest  sig- 
nal-to-noise ratio  in  an  amplifier  the  input 
loading  resistance  should  be  as  high  as  possi- 
ble so  that  the  amount  of  voltage  that  can  be 
developed  from  grid  to  ground  by  the  antenna 
energy  will  be  as  high  as  possible.  The  equi- 
valent noise  resistance  should  be  as  low  as 
possible  so  that  the  noise  generated  by  this 
resistance  will  be  lower  than  that  attributable 
to  the  antenna  and  first  tuned  circuit,  and  the 
losses  in  the  first  tuned  circuit  should  be  as 
low  as  possible. 

The  absolute  sensitivity  of  receivers  has 
been  designated  in  recent  years  in  government 
and  commercial  work  by  an  arbitrary  dimension- 
less number  known  as  "noise  factor”  or  N. 
The  noise  factor  is  the  ratio  of  noise  output 
of  a "perfect”  receiver  having  a given  amount 
of  gain  with  a dummy  antenna  matched  to  its 
input,  to  the  noise  output  of  the  receiver  under 
measurement  having  the  same  amount  of  gain 
with  the  dummy  antenna  matched  to  its  input. 
Although  a perfect  receiver  is  not  a physically 
realizable  thing,  the  noise  factor  of  a receiver 
under  measurement  can  be  determined  by  cal- 
culation from  the  amount  of  additional  noise 
(from  a temperature-limited  diode  or  other  cali- 
brated noise  generator)  required  to  increase 
the  noise  power  output  of  a receiver  by  a pre- 
determined amount. 

Tube  Input  As  has  been  mentioned  in  a pre- 
Looding  vious  paragraph,  greatest  gain 

in  a receiver  is  obtained  when 
the  antenna  is  matched,  through  the  r-f  cou- 
pling transformer,  to  the  input  resistance  of 
the  r-f  tube.  However,  the  higher  the  ratio  of 
tube  input  resistance  to  equivalent  noise  re- 
sistance of  the  tube  the  higher  will  be  the  sig- 
nal-to-noise  ratio  of  the  stage— and  of  course, 
the  better  will  be  the  noise  factor  of  the  over- 
all receiver.  The  input  resistance  of  a tube 
is  very  high  at  frequencies  in  the  broadcast 
band  and  gradually  decreases  as  the  frequency 
increases.  Tube  input  resistance  on  conven- 
tional tube  types  begins  to  become  an  import- 
ant factor  at  frequencies  of  about  25  Me.  and 
above.  At  frequencies  above  about  100  Me.  the 
use  of  conventional  tube  types  becomes  im- 
practicable since  the  input  resistance  of  the 
tube  has  become  so  much  lower  than  the  equi- 
valent noise  resistance  that  it  is  impossible 
to  attain  reasonable  signal-to-noise  ratio  on 
any  but  very  strong  signals.  Hence,  special 
v-h-f  tube  types  such  as  the  6AK5,  6AG5,  and 
6CB6  must  be  used. 

The  lowering  of  the  effective  input  resist- 


HANDBOOK 


R-F  Amplifiers  151 


ance  of  a vacuum  tube  at  higher  frequencies 
is  brought  about  by  a number  of  factors.  The 
first,  and  most  obvious,  is  the  fact  that  the 
dielectric  loss  in  the  internal  insulators,  and 
in  the  base  and  press  of  the  tube  increases 
with  frequency.  The  second  factor  is  due  to 
the  fact  that  a finite  time  is  required  for  an 
electron  to  move  from  the  space  charge  in  the 
vicinity  of  the  cathode,  pass  between  the  grid 
wires,  and  travel  on  to  the  plate.  The  fact  that 
the  electrostatic  effect  of  the  grid  on  the  mov- 
ing electron  acts  over  an  appreciable  portion 
of  a cycle  at  these  high  frequencies  causes  a 
current  flow  in  the  grid  circuit  which  appears 
to  the  input  circuit  feeding  the  grid  as  a re- 
sistance. The  decrease  in  input  resistance  of 
a tube  due  to  electron  transit  time  varies  as 
the  square  of  the  frequency.  The  undesirable 
effects  of  transit  time  can  be  reduced  in  cer- 
tain cases  by  the  use  of  higher  plate  voltages. 
Transit  time  varies  inversely  as  the  square 
root  of  the  applied  plate  voltage. 

Cathode  lead  inductance  is  an  additional 
cause  of  reduced  input  resistance  at  high  fre- 
quencies. This  effect  has  been  reduced  in  cer- 
tain tubes  such  as  the  6SH7  and  the  6AK5  by 
providing  two  cathode  leads  on  the  tube  base. 
One  cathode  lead  should  be  connected  to  the 
input  circuit  of  the  tube  and  the  other  lead 
should  be  connected  to  the  by-pass  capacitor 
for  the  plate  return  of  the  tube. 

The  reader  is  referred  to  the  Radiation  Labo- 
ratory Series,  Volume  23:  "Microwave  Receiv- 
ers” (McGraw-Hill,  publishers)  for  additional 
information  on  noise  factor  and  input  loading 
of  vacuum  tubes. 


8-2  Plate-Circoil 

Considerations 


d)  AMPLIFICATION  AT  RESONANCE  (APPROX>GmU)MQ 


© AMPLIFICATION  AT  RESONANCE(APPRO»GmK 

Q'pQs 

WHERE:  1.  PRI.  AND  SEC.  RESONANT  AT  SAME  FREQUENCY 
2.  K IS  COEFFICIENT  OF  COUPLING 

IF  PRI.  AND  SEC,  Q ARE  APPROXIMATELY  THE  SAME; 

TOTAL  BANDWIDTH  - , , u 
CENTER  FREQUENCY  “ ^ 

MAXIMUM  AMPLITUDE  OCCURS  AT  CRITICAL  COUPLING  - 


Noise  is  generated  in  a vacuum  tube  by  the 
fact  that  the  current  flow  within  the  tube  is  not 
a smooth  flow  but  rather  is  made  up  of  the  con* 
tinuous  arrival  of  particles  (electrons)  at  a 
very  high  rate.  This  shot  effect  is  a source  of 
noise  in  the  tube,  but  its  effect  is  referred 
back  to  the  grid  circuit  of  the  tube  since  it  is 
included  in  the  equivalent  noise  resistance 
discussed  in  the  preceding  paragraphs. 

Plate  Circuit  For  the  purpose  of  this  section. 
Coupling  it  will  be  considered  that  the 

function  of  the  plate  load  cir* 
cuit  of  a tuned  vacuum-tube  amplifier  is  to  de- 
liver energy  to  the  next  stage  with  the  greatest 
efficiency  over  the  required  band  of  frequen- 
cies. Figure  1 shows  three  methods  of  inter- 
stage coupling  for  tuned  r-f  voltage  amplifiers. 
In  figure  lA  omega  {co)  is  In  times  the  reso- 
nant frequency  of  the  circuit  in  the  plate  of 


Figure  1 

Cain  equations  for  pentode  r-f  amplifier 
stages  operating  into  a tuned  load 

the  amplifier  tube,  and  L and  Q are  the  induct- 
ance and  Q of  the  inductor  L.  In  figure  IB  the 
notation  is  the  same  and  M is  the  mutual  in- 
ductance between  the  primary  coil  and  the  sec- 
ondary coil.  In  figure  1C  the  notation  is  again 
the  same  and  k is  the  coefficient  of  coupling 
between  the  two  tuned  circuits.  As  the  co- 
efficient of  coupling  between  the  circuits  is 
increased  the  bandwidth  becomes  greater  but 
the  response  over  the  band  becomes  progres- 
sively more  double-humped.  The  response  over 
the  band  is  the  most  flat  when  the  Q’s  of  pri- 
mary and  secondary  are  approximately  the  same 
and  the  value  of  each  Q is  equal  to  1.75/k. 


152  R-F  Vacuum  Tube  Amplifiers 


the  radio 


Varlable-Mu  Tubes  It  is  common  practice  to 
in  R-F  Stages  control  the  gain  of  a suc- 

cession of  r-f  or  i-f  am- 
plifier stages  by  varying  the  average  bias  on 
their  control  grids.  However,  as  the  bias  is 
raised  above  the  operating  value  on  a conven- 
tional sharp-cutoff  tube  the  tube  becomes  in- 
creasingly non-linear  in  operation  as  cutoff  of 
plate  current  is  approached.  The  effect  of  such 
non-linearity  is  to  cause  cross  modulation  be- 
tween strong  signals  which  appear  on  the  grid 
of  the  tube.  When  a tube  operating  in  such  a 
manner  is  in  one  of  the  first  stages  of  a re- 
ceiver a number  of  signals  are  appearing  on  its 
grid  simultaneously  and  cross  modulation  be- 
tween them  will  take  place.  The  result  of  this 
effect  is  to  produce  a large  number  of  spurious 
signals  in  the  output  of  the  receiver— in  most 


cases  these  signals  will  carry  the  modulation 

of  both  the  carriers  which  have  been  cross 
modulated  to  produce  the  spurious  signal. 

The  undesirable  effect  of  cross  modulation 
can  be  eliminated  in  most  cases  and  greatly 
reduced  in  the  balance  through  the  use  of  a 
variable-mu  tube  in  all  stages  which  have  a-v-c 
voltage  or  other  large  negative  bias  applied  to 
their  grids.  The  variable-mu  tube  has  a char- 
acteristic which  causes  the  cutoff  of  plate  cur- 
rent to  be  gradual  with  an  increase  in  grid 
bias,  and  the  reduction  in  plate  current  is  ac- 
companied by  a decrease  in  the  effective  am- 
plification factor  of  the  tube.  Vatiable-mu  tubes 
ordinarily  have  somewhat  reduced  as  com- 
pared to  a sharp-cutoff  tube  of  the  same  group. 
Hence  the  sharp-cutoff  tube  will  perform  best 
in  stages  to  which  a-v-c  voltage  is  not  applied. 


RADIO-FREQUENCY  POWER  AMPLIFIERS 


All  modern  transmitters  in  the  medium-fre- 
quency range  and  an  increasing  percentage  of 
those  in  the  v-h-f  and  u-h-f  ranges  consist  of 
a comparatively  low-level  source  of  radio-fre- 
quency energy  which  is  multiplied  in  frequency 
and  successively  amplified  to  the  desired  power 
level.  Microwave  transmitters  are  still  predom- 
inately of  the  self-excited  oscillator  type,  but 
when  it  is  possible  to  use  r-f  amplifiers  in 
s-h-f  transmitters  the  flexibility  of  their  ap- 
plication will  be  increased.  The  following  por- 
tion of  this  chapter  will  be  devoted,  however, 
to  the  method  of  operation  and  calculation  of 
operating  characteristics  of  r-f  power  ampli- 
fiers for  operation  in  the  range  of  approximate- 
ly 3.5  to  500  Me. 

8-3  Class  C R-F 

Power  Amplifiers 

The  majority  of  r-f  power  amplifiers  fall  into 
the  Class  C category  since  such  stages  can 
be  made  to  give  the  best  plate  circuit  efficien- 
cy of  any  present  type  of  vacuum-tube  ampli- 
fier. Hence,  the  cost  of  tubes  for  such  a stage 
and  the  cost  of  the  power  to  supply  that  stage 
is  least  for  any  given  power  output.  Neverthe- 
less, the  Class  C amplifier  gives  less  power 
gain  than  either  a Class  A or  Class  B ampli- 
fier under  similar  conditions  since  the  grid  of 
a Class  C stage  must  be  driven  highly  posi- 
tive over  the  portion  of  the  cycle  of  the  excit- 
ing wave  when  the  plate  voltage  on  the  ampli- 
fier is  low,  and  must  be  at  a large  negative 
potential  over  a large  portion  of  the  cycle  so 


that  no  plate  current  will  flow  except  when 
plate  voltage  is  very  low.  This,  in  fact,  is  the 
fundamental  reason  why  the  plate  circuit  effi- 
ciency of  a Class  C amplifier  stage  can  be 
made  high-plate  current  is  cut  off  at  all  times 
except  when  the  plate-to-cathode  voltage  drop 
across  the  tube  is  at  its  lowest  value.  Class 
C amplifiers  almost  invariably  operate  into  a 
tuned  tank  circuit  as  a load,  and  as  a result 
are  used  as  amplifiers  of  a single  frequency 
or  of  a comparatively  narrow  band  of  frequen- 
cies. 

Relofionships  in  Figure  2 shows  the  relation- 
Class  C Stage  ships  between  the  various 
voltages  and  currents  over 
one  cycle  of  the  exciting  grid  voltage  for  a 
Class  C amplifier  stage.  The  notation  given  in 
figure  2 and  in  the  discussion  to  follow  is  the 
same  as  given  at  the  first  of  Chapter  Six  un- 
der ’'Symbols  for  Vacuum-Tube  Parameters.** 
The  various  manufacturers  of  vacuum  tubes 
publish  booklets  listing  in  adequate  detail  al- 
ternative Class  C operating  conditions  for  the 
tubes  which  they  manufacture.  In  addition, 
operating  condition  sheets  for  any  particular 
type  of  vacuum  tube  are  available  for  the  ask- 
ing from  the  different  vacuum-tube  manufac- 
turers, It  is,  nevertheless,  often  desirable  to 
determine  optimum  operating  conditions  for  a 
tube  under  a particular  set  of  circumstances. 
To  assist  in  such  calculations  the  following 
paragr^hs  are  devoted  to  a method  of  calcu- 
lating Class  C operating  conditions  which  is 
moderately  simple  and  yet  sufficiently  accu- 
rate for  all  practical  purposes. 


HANDBOOK 


Class  C R-F  Amplifiers  153 


Figure  2 

Instantaneous  electrode  and  tank  circuit 
voltages  and  currents  for  a Class  C r~f  power 
amplifier 


Calculation  of  Class  Although  Class  C op- 
C Amplifier  Operating  crating  conditions  can 
Characteristics  be  determined  with  the 

aid  of  the  more  conven- 


tional grid  voltage-plate  current  operating 
curves,  the  calculation  is  considerably  sim- 
plified if  the  alternative  "constant-current 

curve"  of  the  tube  in  question  is  used.  This  is 
true  since  the  operating  line  of  a Class  C am- 
plifier is  a straight  line  on  a set  of  constant- 
current  curves.  A set  of  constant-current  curves 
on  the  250TH  tube  with  a sample  load  line 
drawn  thereon  is  shown  in  figure  5. 

In  calculating  and  predicting  the  operation 
of  a vacuum  tube  as  a Class  C radio-frequency 
amplifier,  the  considerations  which  determine 
the  operating  conditions  are  plate  efficiency, 
power  output  required,  maximum  allowable 
plate  and  grid  dissipation,  maximum  allowable 
plate  voltage  and  maximum  allowable  plate 
current.  The  values  chosen  for  these  factors 
will  depend  both  upon  the  demands  of  a par- 
ticular application  and  upon  the  tube  chosen. 

The  plate  and  grid  currents  of  a Class  C 
amplifier  tube  are  periodic  pulses,  the  dura- 
tions of  which  are  always  less  than  180  de- 
grees. For  this  reason  the  average  grid  cur- 
rent, average  plate  current,  power  output,  driv- 
ing power,  etc.,  cannot  be  directly  calculated 
but  must  be  determined  by  a Fourier  analysis 
from  points  selected  at  proper  intervals  along 
the  line  of  operation  as  plotted  upon  the  con- 
stant-current characteristics.  This  may  be  done 
either  analytically  or  graphically.  While  the 
Fourier  analysis  has  the  advantage  of  accu- 
racy, it  also  has  the  disadvantage  of  being 
tedious  and  involved. 

The  approximate  analysis  which  follows 
has  proved  to  be  sufficiently  accurate  for  most 
applications.  This  type  of  analysis  also  has 
the  advantage  of  giving  the  desired  informa- 
tion at  the  first  trial.  The  system  is  direct  in 
giving  the  desired  information  since  the  im- 
portant factors,  power  output,  plate  efficiency, 
and  plate  voltage  are  arbitrarily  selected  at 
the  beginning. 

Method  of  The  first  step  in  the  method  to 
Calculation  be  described  is  to  determine  the 
power  which  must  be  delivered 
by  the  Class  C amplifier.  In  making  this  deter- 
mination it  is  well  to  remember  that  ordinarily 
from  5 to  10  pet  cent  of  the  power  delivered 
by  the  amplifier  tube  or  tubes  will  be  lost  in 
well-designed  tank  and  coupling  circuits  at 
frequencies  below  20  Me.  Above  20  Me.  the 
tank  and  circuit  losses  are  ordinarily  some- 
what above  10  pet  cent. 

The  plate  power  input  necessary  to  produce 
the  desired  output  is  determined  by  the  plate 
efficiency:  = Pout/Np. 

For  most  applications  it  is  desirable  to  oper- 
ate at  the  highest  practicable  efficiency.  High- 
efficiency  operation  usually  requires  less  ex- 
pensive tubes  and  power  supplies,  and  the 


154  R-F  Vacuum  Tube  Amplifiers 


THE  RADIO 


Figure  3 

Relationship  between  the  peak  value  of  the 
fundamental  component  of  riie  tube  plate  cur~ 
rentf  and  average  plate  current;  as  compared 
to  the  ratio  of  the  instantaneous  peak  value 
of  tube  plate  current,  and  average  plate 
current 


RATIO  


Figure  4 

Relationship  between  the  ratio  of  the  peak 
value  of  the  fundamental  component  of  the 
grid  excitation  voltage,  and  the  average  grid 
bias;  as  compared  to  the  ratio  between  in- 
stantaneous peak  grid  current  and  average 
grid  current 


amount  of  artificial  cooling  required  is  fre- 
quently less  than  for  low-efficiency  operation. 
On  the  other  hand,  high-efficiency  operation 
usually  requires  more  driving  power  and  in- 
volves the  use  of  higher  plate  voltages  and 
higher  peak  tube  voltages.  The  better  types 
of  triodes  will  ordinarily  operate  at  a plate 
efficiency  of  75  to  85  pet  cent  at  the  highest 
rated  plate  voltage,  and  at  a plate  efficiency 
of  65  to  75  pet  cent  at  intermediate  values  of 
plate  voltage. 

The  first  determining  factor  in  selecting  a 
tube  or  tubes  for  a particular  application  is 
the  amount  of  plate  dissipation  which  will  be 
required  of  the  stage.  The  total  plate  dissipa- 
tion rating  for  the  tube  or  tubes  to  be  used  in 
the  stage  must  be  equal  to  or  greater  than  that 
calculated  from:  Pp  = Pin  - Pout- 

After  selecting  a tube  or  tubes  to  meet  the 
power  output  and  plate  dissipation  require- 
ments it  becomes  necessary  to  determine  from 
the  tube  characteristics  whether  the  tube  se- 
lected is  capable  of  the  desired  operation  and, 
if  so,  to  determine  the  driving  power,  grid 
bias,  and  grid  dissipation. 

The  complete  procedure  necessary  to  deter- 
mine a set  of  Class  C amplifier  operating  con- 
ditions is  given  in  the  following  steps: 

1.  Select  the  plate  voltage,  power  output, 
and  efficiency. 


2.  Determine  plate  input  from;  Pjn  = 

P OUt/Np. 

3.  Determine  plate  dissipation  from: 

Pp-  Pin  - Pout-  Pp  must  not  exceed 
maximum  rated  plate  dissipation  for  tube 
or  tubes  selected. 

4.  Determine  average  plate  current  from: 

lb  — P in/Pbb- 

5.  Determine  approximate  ip^ax  from: 
fpmax  = 4.9  Ib  for  Np  = 0.85 

Ip  max  — 4.5  Ib  fur  Np  — 0.80 
1pm  ax  — 4.0  Ib  for  Np  = 0.75 
ipmax  = 3.5  Ib  for  Np  = 0.70 

6.  Locate  the  point  on  constant-current 
characteristics  where  the  constant  plate 
current  line  corresponding  to  the  ap- 
proximate ip  max  determined  in  step  5 
crosses  the  line  of  equal  plate  and  grid 
voltages  (diode  line).  Read  ep,„in  at  this 
point.  In  a few  cases  the  lines  of  con- 
stant plate  current  will  inflect  sharply 
upward  before  reaching  the  diode  line. 
In  these  cases  ep„,;a  should  not  be  read 
at  the  diode  line  but  at  the  point  where 
the  plate  current  line  intersects  a line 
drawn  from  the  origin  through  these 
points  of  inflection. 


HANDBOOK 


Constant  Current  Calculations  1 55 


Active  portion  of  the  operating  load  line  for  an  Btmac  250TH  Class  C r-f  power  amplifier, 
showing  first  trial  point  and  the  final  operating  point 


7.  OalculaCC  f'p ni  from;  Epm  “ ^bb  ^pmin' 

8.  Calculate  the  ratio  Ipm/lb  from: 


Ipm  2 Np  Ebb 


9.  From  the  ratio  of  Ipn,/Ib  calculated  in 
step  8 determine  the  ratio  ip^ax/lb  fr°“> 
figure  3. 

10.  Calculate  a new  value  for  ipmax  fro™ 
the  ratio  found  in  step  9. 

ipmax  = (ratio  from  step  9)  Ib 

11.  Read  eg„,p  and  igniaa  from  the  constant- 
current  characteristics  for  the  values  of 
Opmin  and  ip  max  determined  in  steps  6 
and  10. 

12.  Calculate  the  cosine  of  one-half  the 
angle  of  plate  current  flow  from: 


13.  Calculate  the  grid  bias  voltage  from: 


1 

Epc  — X 

1 — cos  dp 


for  tetrodes,  where  /i,2  is  the  grid-screen 
amplification  factor,  and  E^^  is  the  d-c 
screen  voltage. 

14.  Calculate  the  peak  fundamental  grid  ex- 
citation voltage  from: 

Egm  ~ ®gmp  ~ Ejc 

15.  Calculate  the  ratio  Egj,/Ecc  for  the  val- 


156  R-F  Vacuum  Tube  Amplifiers 


THE  RADIO 


ues  of  Epc  and  found  in  steps  13 
and  14. 

16.  Read  igmax/lc  ftom  figure  4 for  the  ratio 
Egm/Ecc  found  in  step  15. 

17.  Calculate  the  average  grid  current  from 
the  ratio  found  in  step  16,  and  the  value 
°f  igmax  found  in  step  11: 

^gm  ax 

I ^ ~ 

Ratio  from  step  16 

18.  Calculate  approximate  grid  driving  pow- 
er from; 

Pd  = 0.9 

19.  Calculate  grid  dissipation  from: 

Pg  = Pd  + Eccic 

Pg  must  not  exceed  the  maximum  rated 
grid  dissipation  for  the  tube  selected. 


Sample  A typical  example  of  a Class  C 

Calculation  amplifier  calculation  is  shown 
in  the  example  below.  Reference 
is  made  to  figures  3,  4 and  5 in  the  calcula- 
tion. 

1.  Desired  power  output-800  watts. 

2.  Desired  plate  voltage— 3500  volts. 
Desired  plate  efficiency-80  pet  cent 
(Np  = 0.80) 

Pin  = 800/0.80  = 1000  watts 

3.  Pp  = 1000  - 800  = 200  watts 

Use  250TH;  max.  Pp  = 250w;  p = 37. 

4.  I),  = 1000/3500  = 0.285  ampere  (285  ma.) 
Max.  Ib  for  250TH  is  350  ma. 

5.  Approximate  ipmax  = 0.285  X 4.5 
= 1.28  ampere 

6-  epmin  = 260  volts  (see  figure  5 first 
trial  point) 

7.  Ep„  = 3500  - 260  = 3240  volts 

8.  Ipm/Ib  = 2X0.80X3500/3240  = 
5600/3240  = 1.73 

9-  ip  m ax/lb  - 4.1  (from  figure  3) 

10.  ip„ax  = 0.285X4.1  = 1.17 

11.  egmp  = 240  volts 
igmax  = 0.430  amperes 

(Both  above  from  final  point  on  figure  5) 


12. 

cos  0p  = 2.32  (1.73  - 

1.57)  = 0.37 

(6ip  = 68.3”) 

1 

13. 

Ecc  = X 

1 - 0.37 

r / 5240  . 

3500 

0.37  ( 240) 

— 

L V 37  J 

37  J 

— “ 240  volts 


14.  Egm  — 240  ( 240)  — 480  volts  grid 

swing 

15.  Eg„/Ecc  = 480/  “ 240  = - 2 
igmax/Ic  - 5.75  (from  figure  4) 

17.  Ic  = 0.430/5.75  =^0.075  amp.  (75  ma. 
grid  current) 

18.  Pd  = 0.9X480X0.075  = 32.5  watts 
driving  power 

19.  Pg  = 32.5  - (-240X0.75)  = 14.5  watts 
grid  dissipation 

Max.  Pg  for  250TH  is  40  watts 

The  power  output  of  any  type  of  r-f  ampli- 
fier is  equal  to: 

Ip  m^p  m/2  — P o 

Ipm  can  be  determined,  of  course,  from  the 
ratio  determined  in  step  8 above  (in  this  type 
of  calculation)  by  multiplying  this  ratio  times 
Ib- 

It  is  frequently  of  importance  to  know  the 
value  of  load  impedance  into  which  a Class 
C amplifier  operating  under  a certain  set  of 
conditions  should  operate.  This  is  simply  Rl= 
Ep„/lpn,.  In  the  case  of  the  operating  condi- 
tions just  determined  for  a 250TH  amplifier 
stage  the  value  of  load  impedance  is: 

Epm  3240 

Rl = 6600  ohms 

Ipm  -495 

Ip  m 

Ip  m ” ^ Ib 

Ib 

Q of  Amplifier  In  order  to  obtain  good  plate 
Tank  Circuit  tank  circuit  tuning  and  low 
radiation  of  harmonics  from 
an  amplifier  it  is  necessary  that  the  plate  tank 
circuit  have  the  correct  Q.  Charts  giving  com- 
promise values  of  Q for  Class  C amplifiers 
are  given  in  the  chapter,  Generation  of  R-F 
Energy.  However,  the  amount  of  inductance 
required  for  a specified  tank  circuit  Q under 
specified  operating  conditions  can  be  calcu- 
lated from  the  following  expression: 


HANDBOOK 


Class  B R-F  Amplifiers  157 


Rl 

ojL  = 

Q 

oj  ~ 2 n X.  operating  frequency 

L = Tank  inductance 

Rl  - Required  tube  load  impedance 
Q = Effective  tank  circuit  Q 
A tank  circuit  Q of  12  to  20  is  recommended 
for  all  normal  conditions.  However,  if  a bal- 
anced push-pull  amplifier  is  employed  the  tank 
receives  two  impulses  per  cycle  and  the  cir- 
cuit Q may  be  lowered  somewhat  from  the 
above  values. 

Quick  Method  of  Tbe  plate  circuit  effi- 

Colculoting Amplifier  ciency  of  a Class  B or 

Plofe  Efficiency  Class  C r-f  amplifier 

can  be  determined  from 
the  following  facts.  The  plate  circuit  efficiency 
of  such  an  amplifier  is  equal  to  the  product  of 
two  factors,  F,,  which  is  equal  to  the  ratio  of 
Rpm  to  Ebb  (F,  = Ep„/Ebb)  and  Fj,  which  is 
proportional  to  the  one-half  angle  of  plate  cur- 
rent flow,  ^p.  A graph  of  Fj  against  both  dp 
and  cos  (9p  is  given  in  figure  6.  Either  <9p  or 
cos  6p  may  be  used  to  determine  Fj.  Cos  6 
may  be  determined  either  from  the  procedure 
previously  given  for  making  Class  C amplifier 
computations  or  it  may  be  determined  from  the 
following  expression; 

^cc  Ebb 

cos  Op  = - - ■ 

M Egm  Epm 

Example  of  It  is  desired  to  know  the  one-half 
Method  angle  of  plate  current  flow  and 

the  plate  circuit  efficiency  for 
an  812  tube  operating  under  the  following  con- 
ditions which  have  been  assumed  from  inspec- 
tion of  the  data  and  curves  given  in  the  RCA 
Transmitting  Tube  Handbook  HB-3: 

Ebb  ~ 1100  volts 
Ecc  = —40  volts 
29 

Egnj  = 120  volts 
Epm  = 1000  volts 

2-  Rl  = Rpm/Ebb  = 0.91 

- 29  X 40  + 1100 

3.  cos  Op  = r: 

29  X 120  - 1000 

60 

= 0.025 

2480 

4.  Fj  = 0.79  (by  reference  to  figure  6) 


Figure  6 

Relationship  between  Factor  F2  and  the 
balFangle  of  plate  current  flow  in  an  ampli- 
fier with  sine-wave  input  and  output  voltage, 
operating  at  a grid-bias  voltage  greater  than 
cut-off 


5.  Np=  F.  X Fj  = 0.91  X 0.79  = 0.72 
(72  per  cent  efficiency) 

F,  could  be  called  the  plate-voltage-swing 
efficiency  factor,  and  Fj  can  be  called  the 
operating-angle  efficiency  factor  or  the  maxi- 
mum possible  efficiency  of  any  stage  running 
with  that  value  of  half-angle  of  plate  current 
flow. 

Np  is,  of  course,  only  the  ratio  between 
power  output  and  power  input.  If  it  is  desired 
to  determine  the  power  input,  exciting  power, 
and  grid  current  of  the  stage,  these  can  be  ob- 
tained through  the  use  of  steps  7,  8,  9,  and  10 
of  the  previously  given  method  for  power  in- 
put and  output;  and  knowing  that  igmax 
0.095  ampere  the  grid  circuit  conditions  can 
be  determined  through  the  use  of  steps  15,  16, 
17,  18  and  19. 

8-4  Class  B Radio 

Frequency  Power  Amplifiers 

Radio  frequency  power  amplifiers  operating 
under  Class  B conditions  of  grid  bias  and  ex- 
citation voltage  are  used  in  two  general  types 
of  applications  in  transmitters.  The  first  gen- 
eral application  is  as  a buffer  amplifier  stage 
where  it  is  desired  to  obtain  a high  value  of 
power  amplification  in  a particular  stage.  A 
particular  tube  type  operated  with  a given 
plate  voltage  will  be  capable  of  somewhat 
greater  output  for  a certain  amount  of  excita- 
tion power  when  operated  as  a Class  B ampli- 


158  R-F  Vacuum  Tube  Amplifiers 


THE  RADIO 


Figure  7 

AVERAGE  PLATE  CHARACTERISTICS  OF  813  TUBE 


fier  than  when  operated  as  a Class  C ampli- 
fier. 

Calculation  of  Calculation  of  the  operating 
Operating  conditions  for  this  type  of 

Characteristics  Class  B t-f  amplifier  can  be 
carried  out  in  a manner  simi- 
lar to  that  described  in  the  previous  para- 
graphs, except  that  the  grid  bias  voltage  is  set 
on  the  tube  before  calculation  at  the  value: 
Ecc  = ~ Ebb/ ft.  Since  the  grid  bias  is  set  at 
cutoff  the  one-half  angle  of  plate  current  flow 
is  90°;  hence  cos  0p  is  fixed  at  0.00.  The 
plate  circuit  efficiency  for  a Class  D r-f  am- 
plifier operated  in  this  manner  can  be  deter- 
mined in  the  following  manner: 


The  "Class  B The  second  type  of  Class  B 

Linear"  r-f  amplifier  is  the  so-called 

Class  B linear  amplifier  which 
is  often  used  in  transmitters  for  the  amplifica- 
tion of  a single- sideband  signal  or  a conven- 
tional amplitude-modulated  wave.  Calculation 
of  operating  conditions  may  be  carried  out  in 
a manner  similar  to  that  previously  described 
with  the  following  exceptions:  The  first  trial 
operating  point  is  chosen  on  the  basis  of  the 
100  per  cent  positive  modulation  peak  of  the 
modulated  exciting  wave.  The  plate  circuit 
and  grid  peak  voltages  and  currents  can  then 
be  determined  and  the  power  input  and  output 


calculated.  Then,  with  the  exciting  voltage 
reduced  to  one-half  for  the  no-modulation  con- 
dition of  the  exciting  wave,  and  with  the  same 
value  of  load  resistance  reflected  on  the  tube, 
the  plate  input  and  plate  efficiency  will  drop 
to  approximately  one-half  the  values  at  the 
100  per  cent  positive  modulation  peak  and  the 
power  output  of  the  stage  will  drop  to  one- 
fourth  the  peak-modulation  value.  On  the  nega- 
tive modulation  peak  the  input,  efficiency,  and 
output  all  drop  to  zero. 

In  general,  the  proper  plate  voltage,  bias 
voltage,  load  resistance  and  power  output 
listed  in  the  tube  tables  for  Class  B audio 
work  will  also  apply  to  Class  B linear  r-f  ap- 
plication. 

Calculation  of  0 per-  Eigure  7 illustrates 

ating  Porometers  for  a the  characteristic 

Class  B Linear  Amplifier  curves  for  an  813 
tube.  Assume  the 
plate  supply  to  be  2000  volts,  and  the  screen 
supply  to  be  400  volts.  To  determine  the  oper- 
ating parameters  of  this  tube  as  a Class  B lin- 
ear r-f  amplifier,  the  following  steps  should 
be  taken: 

1,  The  grid  bias  is  chosen  so  that  the  rest- 
ing plate  current  will  produce  approxi- 
mately 1/3  of  the  maximum  plate  dissi- 
pation of  the  tube.  The  maximum  dissi- 
pation of  the  813  is  125  watts,  so  the 
bias  is  set  to  allow  one-third  of  this 
value,  or  42  watts  of  resting  dissipation. 
At  a plate  potential  of  2000  volts,  a 


HANDBOOK 


Linear  Amplifier  Parameters  159 


plate  current  of  21  milliamperes  will 
produce  this  figure.  Referring  to  figure 
7,  a grid  bias  of  —45  volts  is  approxi- 
mately correct. 

2.  A practical  Class  B linear  r-f  amplifier 
runs  at  an  efficiency  of  about  66%  at  full 
output,  the  efficiency  dropping  to  about 
33%  with  an  unmodulated  exciting  sig- 
nal. In  the  case  of  single- sideband  sup- 
pressed carrier  excitation,  a no-excita- 
tion condition  is  substituted  for  the  un- 
modulated excitation  case,  and  the  lin- 
ear amplifier  runs  at  the  resting  or  qui- 
escent input  of  42  watts  with  no  exciting 
signal.  The  peak  allowable  power  input 
to  the  813  is: 

Input  Peak  Power  (Wp)  = 

(watt;s) 

Plate  Dissipation  X 100 

(100  ~ % plate  efficiency) 

125 

X 100  = 379  watts 

33 

3.  The  maximum  signal  plate  current  is: 

Wp  379 

ipmai  = — = = 0.189  ampere 

Ep  2000 

4.  The  plate  current  flow  of  the  linear  am- 
plifier is  1809,  and  the  plate  current 
pulses  have  a peak  of  3.14  times  the 
maximum  signal  current: 

3.14  X 0.189  = 0.595  ampere 


5.  Referring  to  figure  7,  a current  of  0.605 
ampere  (Point  A)  will  flow  at  a positive 
grid  potential  of  60  volts  and  a minimum 
plate  potential  of  420  volts.  The  grid  is 
biased  at  ~45  volts,  so  a peak  t-f  grid 
voltage  of  60  + 45  volts  = 105  volts  is  re- 
quired. 


PLATE  VOLTS  El- 


Figure  8 

Eg,  VS.  Ep  CHARACTERISTICS  OF  813 
TUBE 


8.  The  plate  load  resistance  is: 

Rl 


O.Sip 


0.5  X .189 
= 6000  ohms 


9.  If  a loaded  plate  tank  circuit  Q of  12  is 
desired,  the  reactance  of  the  plate  tank 
capacitor  at  the  resonant  frequency 
should  be: 


Rl  6000 

Reactance  (ohms)  = = = 500  ohms 

Q 12 


10.  For  an  operating  frequency  of  4.0  Me., 
the  effective  resonant  capacity  is: 


C = 


lOS 


6.28  X 4.0  X 500 


- 80  ppfd. 


6.  The  grid  driving  power  required  for  the 
Class  B linear  stage  maybe  found  by  the 
aid  of  figure  8.  It  is  one-quarter  the  pro- 
duct of  the  peak  grid  current  times  the 
peak  grid  voltage: 

0.02  X 105 

Rg  - 0-53  watt 

4 

7.  The  single  tone  power  output  of  the  813 
stage  is: 

Rp  — 78.5  {Ep  — epmin)  X Ip 
Pp  = 78.5  (2000  - 420)x  .189  = 235  watts 


11.  The  inductance  required  to  resonate  at 
4.0  Me.  with  this  value  of  capacity  is: 

500 

L = 19.9  microhenries 

6.28  X 4.0 


Grid  Circuit  1.  The  maximum  positive  grid 
Considerations  potential  is  60  volts,  and 

the  peak  r-f  grid  voltage  is 
105  volts.  Required  driving  power  is  0.53  watt. 
The  equivalent  grid  resistance  of  this  stage  is: 


160  R-F  Vacuum  Tube  Amplifiers 


THE  RADIO 


(e^)^  105^ 

Rg=  = = 

2XPg  2X0.53 

10,400  ohms 

2.  As  in  the  case  of  the  Class  B audio  am- 
plifier the  grid  resistance  of  the  linear 
amplifier  varies  from  infinity  to  a low 
value  when  maximum  grid  current  is 
drawn.  To  decrease  the  effect  of  this 
resistance  excursion,  a swamping  resis- 
tor should  be  placed  across  the  grid  tank 
circuit.  The  value  of  the  resistor  should 
be  dropped  until  a shortage  of  driving 
power  begins  to  be  noticed.  For  this  ex- 
ample, a resistor  of  3,000  ohms  is  used. 
The  grid  circuit  load  for  no  grid  current 
is  now  3,000  ohms  instead  of  infinity, 
and  drops  to  2400  ohms  when  maximum 
grid  current  is  drawn. 

3.  A circuit  Q of  15  is  chosen  for  the  grid 
tank.  The  capacitive  reactance  required 
is: 

2400 

Xq  = = l60  ohms 

15 

4.  At  4.0  Me.  the  effective  capacity  is: 

10^ 

C=  - 248  ppfd. 

6.28  X 4 X 154 

5.  The  inductive  reactance  required  fo  reso- 
nate the  grid  circuit  at  4.0  Me.  is: 

160 

L = = 6.4  microhenries 

6.28x4.0 

6.  By  substituting  the  loaded  grid  resist- 
ance figure  in  the  formula  in  the  first 
paragraph,  the  grid  driving  power  is  now 
found  to  be  approximately  2.3  watts. 

Screen  Circuit  By  reference  to  the  plate 
Considerations  characteristic  curve  of  the 
813  tube,  it  can  be  seen  that 
at  a minimum  plate  potential  of  500  volts,  arid 
a maximum  plate  current  of  0.6  ampere,  the 
screen  current  will  be  approximately  30  milli- 
amperes,  dropping  to  one  or  two  milliamperes 
in  the  quiescent  state.  It  is  necessary  to  use 
a well-regulated  screen  supply  to  hold  the 
screen  voltage  at  the  correct  potential  over 
this  range  of  current  excursion.  The  use  of  an 
electronic  regulated  screen  supply  is  recom- 
mended. 


8-5  Special  R-F  Power 

Amplifier  Circuits 

The  r-f  power  amplifier  discussions  of  Sec- 
tions 8-4  and  8-5  have  been  based  on  the  as- 
sumption that  a conventional  grounded-cathode 
or  cathode-return  type  of  amplifier  was  in  ques- 
tion. It  is  possible,  however,  as  in  the  case  of 
a-f  and  low-level  r-f  amplifiers  to  use  circuits 
in  which  electrodes  other  than  the  cathode  are 
returned  to  ground  insofar  as  the  signal  poten- 
tial is  concerned.  Both  the  plate-return  or 
cathode-follower  amplifier  and  the  grid-return 
or  grounded-grid  amplifier  are  effective  in  cer- 
tain circuit  applications  as  tuned  r-f  power 
amplifiers. 

Disadvantages  of  An  undesirable  aspect  of 

Grounded-Cathode  the  operation  of  cathode- 

Amplifiers  return  r-f  power  amplifiers 

using  triode  tubes  is  that 
such  amplifiers  must  be  neutralized.  Princi- 
ples and  methods  of  neutralizing  r-f  power  am- 
plifiers are  discussed  in  the  chapter  Genera- 
tion of  R-F  Energy,  As  the  frequency  of  opera- 
tion of  an  amplifier  is  increased  the  stage  be- 
comes more  and  more  difficult  to  neutralize 
due  to  inductance  in  the  grid  and  plate  leads 
of  the  tubes  and  in  the  leads  to  the  neutraliz- 
ing capacitors.  In  other  words  the  bandwidth 
of  neutralization  decreases  as  the  frequency 
is  increased.  In  addition  the  very  presence  of 
the  neutralizing  capacitors  adds  additional 
undesirable  capacitive  loading  to  the  grid  and 
plate  tank  circuits  of  the  tube  or  tubes.  To 
look  at  the  problem  in  another  way,  an  ampli- 
fier that  may  be  perfectly  neutralized  at  a fre- 
quency of  30  Me.  may  be  completely  out  of 
neutralization  at  a frequency  of  120  Me.  There- 
fore, if  there  are  circuits  in  both  the  grid  and 
plate  circuits  which  offer  appreciable  imped- 
ance at  this  high  frequency  it  is  quite  possi- 
ble that  the  stage  may  develop  a "parasitic 
oscillation”  in  the  vicinity  of  120  Me. 

Grounded-Grid  This  condition  of  restricted- 

R-F  Amplifiers  range  neutralization  of  r-f 

power  amplifiers  can  be  great- 
ly alleviated  through  the  use  of  a cathode-  t 
return  or  grounded-grid  r-f  stage.  The  grounded- 
grid  amplifier  has  the  following  advantages; 

1.  The  output  capacitance  of  a stage  is  re- 
duced to  approximately  one-half  the  value 
which  would  be  obtained  if  the  same  tube 
or  tubes  were  operated  as  a conventional 
neutralized  amplifier. 

2.  The  tendency  toward  parasitic  oscillations 
in  such  a stage  is  greatly  reduced  since 
the  shielding  effect  of  the  control  grid  be- 


HANDBOOK 


Grounded  Grid  Amplifier  16l 


tween  the  filament  and  the  plate  is  effec- 
tive over  a broad  range  of  frequencies. 

3.  The  feedback  capacitance  within  the  stage 
is  the  plate-to-cathode  capacitance  which 
is  ordinarily  very  much  less  than  the  grid- 
to-plate  capacitance.  Hence  neutralization 
is  ordinarily  not  required.  If  neutralization 
is  required  the  neutralizing  capacitors  are 
very  small  in  value  and  are  cross  con- 
nected between  plates  and  cathodes  in  a 
push-pull  stage,  or  between  the  opposite 
end  of  a split  plate  tank  and  the  cathode 
in  a single-ended  stage. 

The  disadvantages  of  a grounded-grid  am- 
plifier are; 

1.  A large  amount  of  excitation  energy  is  re- 
quired. However,  only  the  normal  amount 
of  energy  is  lost  in  the  grid  circuit  of  the 
amplifier  tube;  all  additional  energy  over 
this  amount  is  delivered  to  the  load  cir- 
cuit as  useful  output. 

2.  The  cathode  of  a grounded-grid  amplifier 
stage  is  "hot”  to  r.f.  This  means  that  the 
cathode  must  be  fed  through  a suitable  im- 
pedance from  the  filament  supply,  or  the 
secondary  of  the  filament  transformer  must 
be  of  the  low-capacitance  type  and  ade- 
quately insulated  for  the  r-f  voltage  which 
will  be  present. 

3.  A grounded-gri  d r-f  amplifier  cannot  be 
plate  modulated  100  per  cent  unless  the 
output  of  the  exciting  stage  is  modulated 
also.  Approximately  70  per  cent  modula- 
tion of  the  exciter  stage  as  the  final  stage 
is  being  modulated  100  pet  cent  is  recom- 
mended. However,  the  grounded-grrd  r-f 
amplifier  is  quite  satisfactory  as  a Class 
B linear  r-f  amplifier  for  single  sideband 
or  conventional  amplitude  modulated  waves 
or  as  an  amplifier  for  a straight  c-w  or 
FM  signal. 

Figure  9 shows  a simplified  representation 
of  a grounded-grid  ttiode  r-f  power  amplifier 
stage.  The  relationships  between  input  and 
out  put  power  and  the  peak  fundamental  com- 
ponents of  electrode  voltages  and  currents  are 
given  below  the  drawing.  The  calculation  of 
the  complete  operating  conditions  for  a 
grounded-grid  amplifier  stage  is  somewhat  more 
complex  than  that  for  a conventional  amplifier 
because  the  input  circuit  of  the  tube  is  in 
series  with  the  output  circuit  as  far  as  the 
load  is  concerned.  The  primary  result  of  this 
effect  is,  as  stated  before,  that  considerably 
more  power  is  required  from  the  driver  stage. 
The  normal  power  gain  for  a g-g  stage  is  from 
3 to  15  depending  upon  the  grid  circuit  con- 
ditions chosen  for  the  output  stage.  The  higher 
the  grid  bias  and  grid  swing  required  on  the 


Figure  9 

GROUNDED-GRID  CLASS  B OR  CLASS  C 
AMPLIFIER 

T/ie  equations  in  the  above  figure  give  the 
relationships  between  the  fundamental  com- 
ponents of  grid  and  plate  potential  and  cur- 
rent, and  the  power  input  and  power  output 
of  the  stage.  An  expression  for  the  approxi- 
mate cathode  impedance  is  given 


output  stage,  the  higher  will  be  the  require- 
ment from  the  driver. 

Calculation  of  Operating  It  is  most  convenient 
Conditions  of  Grounded  to  determine  the  op- 
Grid  R-F  Amplifiers  erating  conditions  for 

a Class  B or  Class  C 
grounded-grid  r-f  power  amplifier  in  a two-step 
process.  The  first  step  is  to  determine  the 
plate-circuit  and  grid-circuit  operating  condi- 
tions of  the  tube  as  though  it  were  to  operate 
as  a conventional  cathode-return  amplifier 
stage.  The  second  step  is  then  to  add  in  the 
additional  conditions  imposed  upon  the  oper- 
ating conditions  by  the  fact  that  the  stage  is 
to  operate  as  a grounded-grid  amplifier. 

For  the  first  step  in  the  calculation  the  pro- 
cedure given  in  Section  8-3  is  quite  satisfac- 
tory and  will  be  used  in  the  example  to  follow. 
Suppose  we  take  for  our  example  the  case  of  a 
type  304TL  tube  operating  at  2700  plate  volts 
at  a kilowatt  input.  Following  through  the  pro- 
cedure previously  given: 

1.  Desired  power  output— 850  watts 
Desired  Plate  voltage— 2700  volts 
Desired  plate  efficiency— 85  per  cent 
(Np  = 0.85) 


162  R-F  Vacuum  Tube  Amplifiers 


THE  RADIO 


2.  Pi„  = 850/0.85  = 1000  watts 

3.  Pp  = 1000  ~ 850  = 150  watts 
Type  304TL  chosen;  max.  Pp  = 300 
watts,  fi  = 12. 

4.  Ib  = 1000/2700  = 0.370  ampere 
(370  ma.) 

5.  Approximate  ipmax  - 4.9  X 0.370  = 1.81 
ampere 

0-  Cpn,i„=  140  volts  (from  304TL  con- 
stant-current curves) 

7.  Ep„  = 2700  - 140  = 2560  volts 

8.  Ip,„/Ih  = 2 X 0.85  X 2700/2560  = 1.79 

9.  ipmax/lb  = 4.65  (from  figure  3) 

10.  ipmax  - 4.65  X 0.370  = 1.72  amperes 

^gmp  - 140  volts 
Igmax  = 0.480  amperes 

12.  Cos  6p  = 2.32  (1.79-1.57)  = 0.51 
0p  = 59” 

1 

13.  Epc  = X 

1-0.51 

r / 2560  . 2700, 

= “385  volts 

14.  Ego  = 140-(-385)  = 525  volts 

15.  Ego/E,p  = -1.36 

16.  igmai/Ic  “ approx.  8.25  (extrapolated 
from  figure  4) 

17.  4 = 0.480/8.25  = 0.058  (58  ma.  d-c 

grid  current) 

18.  Pd  = 0.9  X 525  X 0.058  = 27.5  watts 

19.  Pg  = 27.5  -( “385  X 0.058)  = 5.2  watts 
Max.  P g for  304TL  is  50  watts 

We  can  check  the  operating  plate  efficiency 
of  the  stage  by  the  method  described  in  Sec- 
tion 8-4  as  follows; 


Fi  = Epo/Ebb  = 2560/2700  = 0.95 

Fj  for  Op  of  59“  (from  figure  6)  = 0.90 

Np  = F,  X Fj  = 0.95  X 0.90  — Approx. 

0.85  (85  per  cent  plate  efficiency) 

Now,  to  determine  the  operating  conditions 
as  a grounded-grid  amplifier  we  must  also  know 
the  peak  value  of  the  fundamental  components 
of  plate  current.  This  is  simply  equal  to 
flpm/lfa)  Ib>  ot: 

Ipm  - 1-79  X 0.370  = 0.660  amperes  (from  4 
and  8 above) 

The  total  average  power  required  of  the 
driver  (from  figure  9)  is  equal  to  Fgn,Ip„/2 
(since  the  grid  is  grounded  and  the  grid  swing 
appears  also  as  cathode  swing)  plus  Pd  which 
is  27.5  watts  from  18  above.  The  total  is: 

525  X 0.660 

Total  drive  = =:  172.5  watts 

2 

plus  27.5  watts  or  200  watts 

Therefore  the  total  power  output  of  the  stage 
is  equal  to  850  watts  (contributed  by  the 
304TL)  plus  172.5  watts  (contributed  by  the 
driver)  or  1022.5  watts.  The  cathode  driving 
impedance  of  the  304TL  (again  referring  to 
figure  7)  is  approximately: 

Z|(  = 525/(0.660  + 0.116)  = approximately  675 

ohms. 


Plate-Return  or  Circuit  diagram,  elec- 

Cothode-Follower  R-F  trode potentials  and  cur- 
Power  Amplifier  rents,  and  operating 

conditions  for  a cath- 
ode-follower r-f  power  amplifier  are  given  in 
figure  10.  This  circuit  can  be  used,  in  addi- 
tion to  the  grounded-grid  circuit  just  dis- 
cussed, as  an  r-f  amplifier  with  a ttiode  tube 
and  no  additional  neutralization  circuit.  How- 
ever, the  circuit  will  oscillate  if  the  imped- 
ance from  cathode  to  ground  is  allowed  to  be- 
come capacitive  rather  than  inductive  or  re- 
sistive with  respect  to  the  operating  frequen- 
cy. The  circuit  is  not  recommended  except  for 
v-h-f  or  u-h-f  work  with  coaxial  lines  as  tuned 
circuits  since  the  peak  grid  swing  required 
on  the  r-f  amplifier  stage  is  approximately 
equal  to  the  plate  voltage  on  the  amplifier 
tube  if  high-efficiency  operation  is  desired. 
This  means,  of  course,  that  the  grid  tank  must 
be  able  to  withstand  slightly  more  peak  volt- 
age than  the  plate  tank.  Such  a stage  may  not 
be  plate  modulated  unless  the  driver  stage  is 
modulated  the  same  percentage  as  the  final 
amplifier.  However,  such  a stage  may  be  used 
as  an  amplifier  or  modulated  waves  (Class  B 
linear)  or  as  a c-w  or  FM  amplifier. 


HANDBOOK 


304TL  G-G  Amplifier  163 


POWER  OUTPUT  TO  LOAD  = 


£pM  (iPM  + Iqm) 


POWER  DELIVERED  BY  OUTPUT  TUBE  = 

Epm  Igm 


= Epm  IPM 


POWER  FROM  DRIVER  TO  LOAD  = • 


TOTAL  POWER  FROM  DRIVER 


Eqm  Igm  _ fEpM  + Qgmp)  Igm 
" 2 “ Z 

= APPflOX.  ^EPM+e^MP)  |.»  L 


ASSUMING  1gm=  i.a  Ic 


circuit.  If  a conventional  filament  transformer 
is  to  be  used  the  cathode  tank  coil  may  con- 
sist of  two  parallel  heavy  conductors  (to  carry 
the  high  filament  current)  by-passed  at  both 
the  ground  end  and  at  the  tube  socket.  The 
tuning  capacitor  is  then  placed  between  fila- 
ment and  ground.  It  is  possible  in  certain  cases 
to  use  two  r-f  chokes  of  special  design  to  feed 
the  filament  current  to  the  tubes,  with  a con- 
ventional tank  circuit  between  filament  and 
ground.  Coaxial  lines  also  may  be  used  to 
serve  both  as  cathode  tank  and  filament  feed 
to  the  tubes  for  v-h-f  and  u-h-f  work. 


8-6  A Grounded-Grid 

304TL  Amplifier 


POWER  ABSORBED  BY  OUTPUT  TUBE  GRID  AND  BIAS  SUPPLY- 

a APPROX.  0.9  (Ecc  6gmp)  Ic 

Zci  = = APPBOX, 

^ Igm  1.8  I c 

Figure  10 

CATHODE-FOLLOWER  R-F  POWER 
AMPLIFIER 

Showing  the  rtlafionships  between  the  tube 
potentials  and  currents  and  the  input  and 
output  power  of  the  stage.  The  approximate 
grid  impedance  also  Is  given. 


The  design  of  such  an  amplifier  stage  is 
essentially  the  same  as  the  design  of  a 
grounded-grid  amplifier  stage  as  far  as  the 
first  step  is  concerned.  Then,  for  the  second 
step  the  operating  conditions  given  in  figure 
10  are  applied  to  the  data  obtained  in  the  first 
step.  As  an  example,  take  the  304TL  stage 
previously  described.  The  total  power  required 
of  the  driver  will  be  (from  figure  10)  approxi- 
mately (2700X0. 58X1. 8)/2  or  141  watts.  Of 
this  141  watts  27.5  watts  (as  before)  will  be 
lost  as  grid  dissipation  and  bias  loss  and  the 
balance  of  113.5  watts  will  appear  as  output. 
The  total  output  of  the  stage  will  then  be  ap- 
proximately 963  watts. 

Cathode  Tank  for  The  cathode  tank  circuit 
G-GorC-F  for  either  a grounded-grid 

Power  Amplifier  or  cathode-follower  r-f 

power  amplifier  may  be  a 
conventional  tank  circuit  if  the  filament  trans- 
former for  the  stage  is  of  the  low-capacitance 
high-voltage  type.  Conventional  filament  trans- 
formers, however,  will  not  operate  with  the 
high  values  of  r-f  voltage  present  in  such  a 


The  304TL  tube  is  capable  of  operating  as 
an  r-f  amplifier  of  the  conventional  type  with 
the  full  kilowatt  input  permitted  amateur  sta- 
tions. The  tube  is  characterized  by  an  enor- 
mous reserve  of  filament  emission,  resulting 
from  the  fact  that  about  130  watts  is  required 
merely  to  light  the  four  filaments.  Where  the 
heavy  filament  drain,  plus  the  low  amplifica- 
tion factor  of  about  12,  does  not  impose  hard- 
ship in  the  design  of  the  transmitter,  the  304TL 
is  quite  satisfactory  for  amateur  service. 

The  304TL  offers  an  additional  feature  in 
that  its  plate-to-cathode  capacitance  is  very 
low  febout  0.6  lifild,)  lor  a tube  of  its  size 
and  power  handling  capabilities.  This  feature 
permits  the  tube  to  be  operated,  without  neu- 
tralization, as  a grounded-grid  r-f  power  am- 
plifier. 

Characteristics  of  Although  the  excitation 

Grounded-Grid  power  of  the  tube  is  only 

Operation  27.5  watts  (which  would  be 

the  actual  driving  power 
required  if  the  tube  were  operating  as  a neu- 
tralized amplifier)  the  driving  power  required 
into  the  cathode  circuit  from  the  exciter  is 
about  200  watts.  The  extra  170  watts  or  so  is 
not  lost,  however,  since  it  appears  directly  in 
the  output  of  the  amplifier  as  additional  ener- 
gy. Thus  while  the  304TL  itself  will  deliver 
about  850  watts  to  the  load  circuit,  the  extra 
170  watts  supplied  by  the  driver  over  and  above 
the  27.5  watts  required  to  excite  the  304TL 
appears  added  to  the  850-watt  output  of  the 
304TL.  Thus  the  total  output  of  the  stage 
would  be  about  1020  watts,  even  though  the 
d-c  input  to  the  304TL  was  only  one  kilowatt. 
Nevertheless,  the  tube  itself  operates  only  at 
its  normal  plate- circuit  efficiency,  the  extra 
power  output  coming  directly  from  the  added 
excitation  power  taken  from  the  driver. 


164  R-F  Vacuum  Tube  Amplifiers 


THE  RADIO 


304-TL  L4 


Figure  11 

SCHEMATIC  OF  THE 
304TL  G-G  AMPLIFIER 

L|— None  required  for  3.5  Me.,  since  cathode  coil  L2 
tunes  to  3.5  Me. 

7 Me.— 20  turns  no.  14  enom.,  1!4'*  dia.  by  3'*  long. 
14  Me. -10  turns  no.  8 bore,  1'4**  dla.  by  3**  long. 
L^^Two  parallel  lengths  of  no.  12  enam.  elose*wound 
to  fill  Notionol  type  XR-10A  coil  form. 

25  turns  of  the  two  wites  in  parallel. 

L3— S'turn  link  of  no.  16  rubber  and  eotton  covered  wire. 
L4— 3.5  Me.:  18  turns  no.  12  enam.;  Not.  XR-10A  form 
wound  full,  4 t.  link 

7 Me.:  10  turns  no.  12  enam.;  Nat.  XR'IOA  form,  3 t. 

link 

14  Me.:  4 turns  no.  8 bare;  Nat.  XR-10A  form,  2 t. 

link 

RFC— 800.ma.  r*f  choke  (National  R'175) 


Since  15  to  20  per  cent  of  the  output  from  a 

grounded-grid  r-f  stage  passes  directly  through 
it  from  the  driver,  it  is  obvious  that  conven- 
tional plate  modulation  of  the  output  stage 
only  will  give  incomplete  modulation.  In  fact, 
if  the  plate  voltage  is  completely  removed  from 
the  g-g  (grounded-grid)  stage  and  the  plate  re  - 
turn  is  grounded,  the  energy  from  the  driver 
will  pass  directly  through  the  tube  and  appear 
in  the  output  circuit.  So  it  becomes  necessary 
to  apply  modulation  to  the  driver  tube  simul- 
taneously to  that  which  is  applied  to  the  plate 
of  the  g-g  stage.  It  is  normal  practice  in  com- 
mercial application  to  modulate  the  driver 
stage  about  60  per  cent  as  much  as  the  g-g 
stage. 

For  straight  c-w  operation  no  special  con- 
siderations ate  involved  in  the  use  of  a g-g 
power  amplifier  except  that  it  obviously  is 
necessary  that  the  driver  stage  be  keyed  simul- 
taneously with  the  final  amplifier,  if  the  in- 
stallation is  such  that  the  final  amplifier  is  to 
be  keyed.  With  excitation,  keying  the  combina- 
tion of  the  driver  and  output  stage  will  oper- 
ate quite  normally. 

The  Grounded-Grid  The  g-g  power  amplifier 
Class  B Linear  is  particularly  well  suited 

to  use  as  a class  B lin- 
ear amplifier  to  build  up  the  output  of  a lower 
powered  SSB  transmitter.  Two  advantages  ob- 
tained through  the  use  of  the  g-g  stage  as  a 
class  B linear  amplifier  are  (1)  no  swamping 
is  required  in  the  input  circuit  of  the  g-g  stage, 
and  (2)  nearly  the  full  output  of  the  exciter 
transmitter  appears  in  the  output  of  the  g-g 
stage,  being  added  to  the  output  of  the  g-g 
amplifier. 

Since  operating  impedances  are  lowered 
when  a stage  is  operated  as  a Class  B linear, 
rather  high  values  of  tank  capacitance  are  re- 
quired for  a satisfactory  operating  circuit  Q. 
The  load  impedance  presented  to  one  304TL 
with  the  operating  conditions  specified  above 
is  about  2500  ohms;  hence  the  plate  tank  cir- 
cuit capacitance  for  an  operating  Q of  15 
should  be  about  180  ohms.  This  represents  an 
operating  plate  tank  capacitance  of  about  240 
fififd.  for  4 Me.,  and  proportionately  smaller 
capacitances  for  the  higher  frequency  bands. 
The  cathode  impedance  for  one  304TL  under 
the  operating  conditions  specified  is  about 
700  ohms.  Thus,  for  an  operating  cathode-cir- 
cuit Q of  10  the  cathode  tank  capacitance 


Figure  12 

REAR  OF  THE  GROUNDED-GRID 
AMPLIFIER 


HANDBOOK 


304TL  G-G  Amplifier  165 


should  have  a reactance  of  70  ohms.  This 
value  of  reactance  is  represented  by  about  550 
ppfd.  at  4 Me.,  and  proportionately  smaller  ca- 
pacitances for  higher  frequencies.  Since  the 
peak  cathode  voltage  is  only  about  400,  quite 
small  spacing  in  the  cathode  tank  capacitor 
can  be  used. 

Figure  11  illustrates  a practical  grounded- 
grid  amplifier,  using  a single  304TL  tube.  A 
bifilar  wound  coil  is  used  to  feed  the  filament 
voltage  to  the  tube.  Shunting  inductors  are 
used  to  tune  the  filament  circuit  to  higher  fre- 
quencies. When  tuning  the  g-g  stage,  it  must 
be  remembered  that  the  input  tuned  circuit  is 
effectively  in  series  with  the  output  circuit  as 
far  as  the  output  energy  is  concerned.  Thus, 
it  is  not  possible  to  apply  full  excitation  power 
to  the  stage  unless  plate  current  is  also  flow- 
ing. If  full  excitation  is  applied  in  the  absence 
of  plate  voltage,  the  grid  of  the  output  tube 
would  be  damaged  from  excessive  excitation. 
Construction  of  the  amplifier  is  illustrated  in 
figures  12  and  13. 

A 3000- volt  plate  supply  is  used  which 
should  have  good  regulation.  An  output  filter 
condenser  of  at  least  10  microfarads  should  be 
employed  for  Class  B operation.  Class  B oper- 
ating bias  is  —260  volts,  obtained  from  a well 
regulated  bias  supply. 

For  operation  in  TV  areas,  additional  lead 
filtering  is  required,  as  well  as  complete 
screening  of  the  amplifier. 


Figure  13 

UNDERSIDE  OF  THE  GROUNDED-GRID 
AMPLIFIER 


8-7  Class  ABl  Radio  Frequency 
Power  Amplifiers 

Class  ABl  r-f  amplifiers  operate  under 
such  conditions  of  bias  and  excitation  that 
grid  current  does  not  flow  over  any  portion  of 
the  input  cycle.  Tliis  is  desirable,  since 
distortion  caused  by  grid  current  loading  is 
absent,  and  also  because  the  stage  is  capable 
of  high  power  gain.  Stage  efficiency  is  about 
58%  when  a plate  current  operating  angle  of 
210^  is  chosen,  as  compared  to  62%  for  Class 
B operation. 

The  level  of  static  (quiescent)  plate  current 
for  lowest  distortion  is  quite  critical  for 
Class  ABl  tetrode  operation.  This  value  is 
determined  by  the  tube  characteristics,  and  is 
not  greatly  affected  by  the  circuit  parameters 
or  operating  voltages.  The  maximum  d-c  plate 
potential  is  therefore  limited  by  the  static 
dissipation  of  the  tube,  since  the  resting  plate 
current  figure  is  fixed.  The  static  plate  current 
of  a tetrode  tube  varies  as  the  3/2  power  of 
the  screen  voltage.  For  example,  raising  the 
screen  voltage  from  300  to  500  volts  will 
double  the  plate  current.  The  optimum  static 
plate  current  for  minimum  distortion  is  also 
doubled,  since  the  shape  of  the  Eg-Ip  curve 
does  not  change. 

In  actual  practice,  somewhat  lower  static 
plate  current  than  optimum  may  be  employed 
without  raising  the  distortion  appreciably, 
and  values  of  static  plate  current  of  0.6  to 
0.8  of  optimum  may  be  safely  used,  depending 
upon  the  amount  of  nonlinearity  that  can  be 
tolerated. 

As  with  the  class  B linear  stage,  the  mini- 
mum plate  voltage  swing  of  the  class  ABl 
amplifier  must  be  kept  above  the  d-c  screen 
potential  to  prevent  operation  in  the  nonlinear 
portion  of  the  characteristic  curve.  A low  value 
of  screen  voltage  allows  greater  r-f  plate 
voltage  swing,  resulting  in  improvement  in 
plate  efficiency  of  the  tube.  A balance  be- 
tween plate  dissipation,  plate  efficiency,  and 
plate  voltage  swing  must  be  achieved  for  best 
linearity  of  the  amplifier. 

The  5-Curve  The  perfect  linear  ampli- 

fier delivers  a signal 
that  is  a replica  of  the  input  signal.  Inspection 
of  the  plate  characteristic  curve  of  a typical 
tube  will  disclose  the  tube  linearity  under 
class  A operating  conditions  (figure  14).  The 
curve  is  usually  of  exponential  shape,  and 
the  signal  distortion  is  held  to  a small  value 
by  operating  the  tube  well  below  its  maximum 
output,  and  centering  operation  over  the  most 
linear  portion  of  the  characteristic  curve. 


166  R -F  Vacuum  Tube  Amplifiers 


THE  RADIO 


Amplifier  operation  is  confined  to 
most  linear  portion  of  characterist ic 
curve. 


The  relationship  between  exciting  voltage 
in  a Class  ABl  amplifier  and  the  t-f  plate 
circuit  voltage  is  shown  in  figure  15.  With  a 
small  value  of  static  plate  current  the  lower 
portion  of  the  line  is  curved.  Maximum  un- 
distorted output  is  limited  by  the  point  on  the 
line  (A)  where  the  instantaneous  plate  voltage 
down  to  the  screen  voltage.  This  ”hook”  in 
the  line  is  caused  by  curtent  divetted  from 
the  plate  to  the  grid  and  screen  elements  of 
the  tube.  The  characteristic  plot  of  the  usual 
linear  amplifier  takes  the  shape  of  an  S-curve. 
The  lower  portion  of  the  curve  is  straightened 
out  by  using  the  proper  value  of  static  plate 
curtent,  and  the  upper  portion  of  the  curve  is 
avoided  by  limiting  minimum  plate  voltage 

swing  to  a point  substantially  above  the  value 
of  the  screen  voltage. 

Operating  Parameters  The  approximate  oper- 

for  the  Class  ABl  ating  parameters  may  be 

Linear  Amplifier  obtained  from  the  Con- 
stant Current  curves 

(Eg-Ep)  or  the  Eg-Ip  curves  of  the  tube  in 
question.  An  operating  load  line  is  first 
approximated.  One  end  of  the  load  line  is 

determined  by  the  d-c  operating  voltage  of  the 
tube,  and  the  requited  static  plate  current. 
As  a starting  point,  let  the  product  of  the  plate 
voltage  and  current  equal  the  plate  dissipation 
of  the  tube.  Assuming  we  have  a 4-400A 
tetrode,  this  end  of  the  load  line  will  fall  on 
point  A (figure  16).  Plate  power  dissipation 
is  360  watts  (3000V  (®  120  ma).  The  opposite 
end  of  the  load  line  will  fall  on  a point  deter- 
mined by  the  minimum  instantaneous  plate  volt- 
age, and  by  the  maximum  instantaneous  plate 
current.  The  minimum  plate  voltage,  for  best 
linearity  should  be  considerably  higher  than 


Figure  15 

LINEARITY  CURVE  OF 
TYPICAL  TETRODE  AMPLIFIER 

At  point  *‘A**  the  instantaneous  plate 
voltage  is  swinging  down  to  the  value 
of  screen  voltage.  At  point  “6**  it  is 
swinging  well  below  the  screen  and  is 
approaching  the  point  where  saturation, 
or  plate  current  limiting  takes  place. 


the  screen  voltage.  In  this  case,  the  screen 
voltage  is  500,  so  the  minimum  plate  voltage 
excursion  should  be  limited  to  600  volts. 
Class  ABl  operation  implies  no  grid  current, 
therefore  the  load  line  cannot  cross  the  Eg=0 
line.  At  the  point  Ep=600,  Eg=0,  the  maximum 
plate  current  is  580  ma  (Point  "B”). 

Each  point  the  load  line  crosses  a grid  volt- 
age axis  may  be  taken  as  a point  for  construc- 
tion of  the  Eg-Ip  curve,  just  as  was  done  in 
figure  22,  chapter  6.  A constructed  curve 
shows  that  the  approximate  static  bias  voltage 
is -74  volts,  which  checks  closely  with  point  A 
of  figure  16.  In  actual  practice,  the  bias  volt- 
age is  set  to  hold  the  actual  dissipation  slightly 
below  the  maximum  figure  of  the  tube. 

The  single  tone  power  output  is: 

Emax -EminX Ipmax.  or  3000— 600  X . 58  = 348  warts 
4 4 

The  plate  current-angle  efficiency  factor  for 
this  class  of  operation  is  0.73,  and  the  actual 
plate  circuit  efficiency  is: 

Nn— £niax~EminXQ.73,  or  3000~600  X 0.73  - 58.4% 
Emax  3000 

The  power  input  to  the  stage  is  therefore 

Po  X 100  or,  348  = 595  watts 
Np  58.4 

The  plate  dissipation  is:  595~348  = 247  watts. 


HANDBOOK 


AB|  R.F.  Power  Amplifiers  167 


AMPLIFIER  ARE  OBTAINED  FROM  CONST ANT-CURRENT 
CURVES. 


It  can  be  seen  that  the  limiting  factor  for 
this  class  of  operation  is  the  static  plate  dissi- 
pation, which  is  quite  a bit  higher  than  the 
operating  dissipation  level.  It  is  possible,  at 
tbe  expense  of  a higher  level  of  distortion,  to 
drop  the  static  plate  dissipation  and  to  increase 
the  screen  voltage  to  obtain  greater  power  out- 
put. If  the  screen  voltage  is  set  at  800,  and 
the  bias  increased  sufficiently  to  drop  the 
static  plate  current  to  90  ma,  the  single  tone 
d-c  plate  current  may  rise  to  300  ma,  for  a 
power  input  of  900  watts.  The  plate  circuit 
efficiency  is  55.6%,  and  the  power  output  is 
500  watts.  Static  plate  dissipation  is  270  watts. 

At  a screen  potential  of  500  volts,  the  maxi- 
mum screen  current  is  less  than  1 ma,  and  under 
certain  loading  conditions  may  be  negative. 
When  the  screen  potential  is  raised  to  800  volts 
maximum  screen  current  is  18  ma.  The  per- 
formance of  the  tube  depends  upon  the  voltage 
fields  set  up  within  the  tube  by  the  cathode, 
control  grid,  screen  grid,  and  plate.  The  quantity 
of  current  flowing  in  the  screen  circuit  is  only 
incidental  to  the  fact  that  the  screen  is  main- 
tained at  a positive  potential  with  respect  to 
the  electron  stream  surrounding  it. 

The  tube  will  perform  as  expected  so  long  as 
the  screen  current,  in  either  direction,  does 
not  create  undesirable  changes  in  the  screen 
voltage,  or  cause  excessive  screen  dissipation. 
Good  regulation  of  the  screen  supply  is  there- 


fore required.  Screen  dissipation  is  highly 
responsive  to  plate  loading  conditions,  and  the 
plate  circuit  should  always  be  adjusted  so  as 
to  keep  the  screen  current  below  the  maximum 
dissipation  level  as  established  by  the  applied 
voltage . 

G-G  Gloss  B Linear  Certain  tetrode  and  pentode 
Tetrode  Amplifier  tubes,  such  as  the  6AG7, 
837,  and  803  perform  well 
as  grounded  grid  class  B linear  amplifiers.  In 
this  configuration  both  grids  and  the  suppressor 
are  grounded,  and  excitation  is  applied  to  the 
cathode  circuit  of  the  tube.  So  connected,  the 
tubes  take  on  characteristics  of  high-mu  triodes. 
No  bias  or  screen  supplies  are  required  for 
this  type  of  operation,  and  reasonably  linear 


6AG7 


Figure  17 

SIMPLE  GROUNDED-GRID 
LINEAR  AMPLIFIER 


168  R-F  Vacuum  Tube  Amplifiers 


THE  RADIO 


Vi 

837 


INPUT 
2 WATTS 
PEAK 


V4  V5 

803  803  500 


.001  — 
5 KV 


Figure  18 

3-STAGE  KILOWATT  LINEAR 
AMPLIFIER  FOR  80  OR  40 
METER  OPERATION 


An  open  frame  filament  transformer  may 
be  used  for  T7.  Cothoc/e  taps  are  ad- 
justed  for  proper  excitation  of  following 
stage . 


operation  can  be  had  with  a very  minimum  of 
circuit  components  (figure  17).  The  input  im- 
pedance of  the  g-g  stage  falls  between  100 
and  250  ohms,  eliminating  the  necessity  of 
swamping  resistors,  even  thougli  considerable 
power  is  drawn  by  the  cathode  circuit  of  the 
g-g  stage. 

Power  gain  of  a g-g  stage  varies  from  ap- 
proximately 20  when  tubes  ‘of  the  6AG7  type 
are  used,  down  to  five  or  six  for  the  837  and 
803  tubes.  One  or  mote  g-g  stages  may  be 
cascaded  to  provide  up  to  a kilowatt  of  power, 
as  illustrated  in  figure  18. 

The  input  and  output  circuits  of  cascaded 
g-g  stages  are  in  series,  and  a variation  in 
load  impedance  of  the  output  stage  reflects 
back  as  a proportional  change  on  the  input 
circuit.  If  the  first  g-g  stage  is  driven  by  a high 
impedance  source,  such  as  a tetrode  amplifier, 
any  change  in  gain  will  automatically  be  com- 
pensated for.  If  the  gain  of  V4-V5  drops,  the 
input  impedance  to  that  stage  will  rise.  This 
change  will  reflect  through  V2-V3  so  that  the 
load  impedance  of  VI  rises.  Since  VI  has  a 
high  internal  impedance  the  output  voltage 
will  rise  when  the  load  impedance  rises.  The 
increased  output  voltage  will  raise  the  output 
voltage  of  each  g-g  stage  so  that  the  overall 
output  is  nearly  up  to  the  initial  value  before 
the  drop  in  gain  of  V4-V5. 

The  tank  circuits,  therefore,  of  all  g-g  stages 
must  be  resonated  with  low  plate  voltage  and 
excitation  applied  to  the  tubes.  Tuning  of  one 
stage  will  affect  the  other  stages,  and  the  in- 
put and  coupling  of  each  stage  must  be  ad- 
justed in  turn  until  the  proper  power  limit  is 
reached. 


Operating  Data  for 
4-400A  Grounded 
Grid  Linear 
Amplifier 


Experiments  have  been 
conducted  by  Collins  Ra- 
dio Co.  on  a grounded  grid 
linear  amplifier  stage 


using  a 4— 400A  tube  for  the  h.f.  region.  The 
operating  characteristics  of  the  amplifier  are 
summarized  in  figure  19*  It  can  be  noted  that 
unusually  low  screen  voltage  is  used  on  the 
tube.  The  use  of  lower  screen  voltage  has  the 
adverse  effect  of  increasing  the  driving  power, 
but  at  the  same  time  the  static  plate  current 
of  the  stage  is  decreased  and  linearity  is  im- 
improved.  For  grounded  grid  operation  of  the 
4-400A,  a screen  voltage  of  300  volts  (filament 
to  screen)  gives  a reasonable  compromise  be- 
tween these  factors. 


OPERATING  DATA  FOR  4-400 A / 4-25.0 A 

GtG.  linear  abi  amplifier 

(S/f^SL£  TONS  ) 

D-C  SCREEN  VOLTAGE 

+-S00 

+ 300 

0-C  PLATE  VOLTAGE 

+ 3000 

+ 3500 

STATIC  PLATE  CURRENT 

60  MA. 

60  MA. 

D-C  GRID  BIAS 

-60  V, 

-59  V. 

PEAK  CATHODE  SWING 

8 7 V. 

11  3 V. 

MINIMUM  PLATE  VOLTAGE 

660  V. 

500  V. 

MAXIMUM  SIGNAL  GRID  CURRENT 

3.  6 MA. 

10  MA. 

MAXIMUM  SIGNAL  SCREEN  CURRENT 

4,  1 MA. 

20  MA. 

MAXIMUM  SIGNAL  PLATE  CURRENT 

1 95  MA. 

267  MA. 

MAXIMUM  SIGNAL  PLATE  DISSIPATION 

235  W. 

235  W. 

STATIC  PLATE  DISSIPATION 

180  W. 

210W. 

GRID  DRIVING  POWER 

0.63  W. 

3.4  W. 

FEEOTHRU  POWER 

6.55  W. 

15.  8 W. 

POWER  OUTPUT  {MAXIMUM) 

350  W. 

700  W. 

POWER  INPUT  (MAXIMUM) 

585  W. 

935  W. 

Figure  19 


HANDBOOK 


169 


TOP  VIEW  OF  A 4-250A  AMPLIFIER 

Pi-neiwork  tetrode  amplifier  may  be  operated  Class  ABi,  Class  B,  or  Class  C by  varying  various  potentials 
applied  to  tube.  Same  general  physical  and  mechanical  des  gn  applies  in  each  case. 


CHAPTER  NINE 


The  Oscilloscope 


The  cathode-ray  oscilloscope  (also  called 
oscillograph)  is  an  instrument  which  permits 
visual  examination  of  various  electrical  phe- 
nomena of  interest  to  the  electronic  engineer. 
Instantaneous  changes  in  voltage,  current  and 
phase  are  observable  if  they  take  place  slow- 
ly enough  for  the  eye  to  follow,  or  if  they  are 
periodic  for  a long  enough  time  so  that  the  eye 
can  obtain  an  impression  from  the  screen  of 
the  cathode-ray  tube.  In  addition,  the  cathode- 
ray  oscilloscope  may  be  used  to  study  any 
variable  (within  the  limits  of  its  frequency 
response  characteristic)  which  can  be  con- 
verted into  electrical  potentials.  This  conver- 
sion is  made  possible  by  the  use  of  some  type 
of  transducer,  such  as  a vibration  pickup  unit, 
pressure  pickup  unit,  photoelectric  cell,  mi- 
crophone, or  a variable  impedance.  The  use 
of  such  a transducer  makes  the  oscilloscope 
a valuable  tool  in  fields  other  than  electronics. 


9-1  A Typical  Cathode-Ray 
Oscilloscope 

For  the  purpose  of  analysis,  the  operation 
of  a simple  oscilloscope  will  be  described. 
The  Du  Mont  type  274-A  unit  is  a fit  instru- 
ment for  such  a description.  The  block  diagram 
of  the  274-A  is  shown  in  figure  1.  The  elec- 
tron beam  of  the  cathode-ray  tube  can  be  moved 
vertically  or  horizontally,  or  the  vertical  and 
horizontal  movements  may  be  combined  to  pro- 
duce composite  patterns  on  the  tube  screen. 
As  shown  in  figure  1,  the  cathode-ray  tube  is 


the  recipient  of  signals  from  two  sources:  the 
vertical  and  horizontal  amplifiers.  The  opera- 
tion of  the  cathode-ray  tube  itself  has  been 
covered  in  Chapter  4;  the  auxiliary  circuits 
pertaining  to  the  cathode-ray  tube  will  be 
coveted  here. 


The  Vertical  The  incoming  signal  which  is 
Amplifier  to  be  examined  is  applied  to 

the  terminals  marked  Vertical 
Input  and  Ground.  The  Vertical  Input  terminal 
is  connected  through  capacitor  C,  (figure  2)  so 
that  the  a-c  component  of  the  input  signal  ap- 
pears across  the  vertical  amplifier  gain  con- 
trol potentiometer,  R,.  Thus  the  magnitude  of 
the  incoming  signal  may  be  controlled  to  pro- 
vide the  desired  deflection  on  the  screen  of 


Figure  1 

BLOCK  DIAGRAM,  TYPE  274-A 
CATHODE-RAY  OSCILLOSCOPE 


170 


Time  Base  Generator  171 


Figure  2 

TYPICAL  AMPLIFIER  SCHEMATIC 


the  cathode-ray  tube.  Also,  as  shown  in  figure 
1,  Sj  has  been  incorporated  to  by-pass  the 
vertical  amplifier  and  capacitively  couple  the 
input  signal  directly  to  the  vertical  deflection 
plate  if  so  desired. 

In  figure  2,  V;  is  a 6AC7  pentode  tube  which 
is  used  as  the  vertical  amplifier.  As  the  sig- 
nal variations  appear  on  the  grid  of  V,,  varia- 
tions in  the  plate  current  of  V,  will  take  place. 
Thus  signal  variations  will  appear  in  opposite 
phase  and  greatly  amplified  across  the  plate 
resistor,  R,.  Capacitor  C,  has  been  added  a- 
cross  Rj  in  the  cathode  circuit  of  Vj  to  flatten 
the  frequency  response  of  the  amplifier  at  the 
high  frequencies.  This  capacitor  because  of 
its  low  value  has  very  little  effect  at  low  in- 
put frequencies,  but  operates  more  effectively 
as  the  frequency  of  the  signal  increases.  The 
amplified  signal  delivered  by  V,  is  now  ap- 
plied through  the  second  half  of  switch  Sj  and 
capacitor  C4  to  the  free  vertical  deflection 
plate  of  the  cathode-ray  tube  (figure  3). 


The  Horizontal  The  circuit  of  the  horizontal 
Amplifier  amplifier  and  the  circuit  of 

the  vertical  amplifier,  de- 
scribed in  the  above  paragraph,  are  similar.  A 
switch  in  the  input  circuit  makes  provision  for 
the  input  from  the  Horizontal  Input  terminals 
to  be  capacitively  coupled  to  the  grid  of  the 
horizontal  amplifier  or  to  the  free  horizontal 
deflection  plate  thus  by-passing  the  amplifier, 
or  for  the  output  of  the  sweep  generator  to  be 
capacitively  coupled  to  the  amplifier,  as  shown 
in  figure  1. 

The  Time  Base  Investigation  of  electrical 
Generator  wave  forms  by  the  use  of  a 

cathode-ray  tube  frequently 
requites  that  some  means  be  readily  available 
to  determine  the  variation  in  these  wave  forms 
with  respect  to  time.  When  such  a time  base 
is  required,  the  patterns  presented  on  the  cath- 
ode-ray  tube  screen  show  the  variation  in  am- 
plitude of  the  input  signal  with  respect  to 


SCHEMATIC  OF  CATHODE-RAY  TUBE 
CIRCUITS 

A SBPIA  cathode»ray  tube  is  used  in  this 
instrument.  As  shown,  the  necessary  pofen- 
tiais  for  operating  this  tube  are  obtained 
from  a voltage  divider  made  up  of  resistors 
1^21  through  1^26  inclusive.  The  intensity  of 
the  beam  is  adjusted  by  moving  the  contact 
on  This  adjusts  the  potential  on  the 

cathode  more  or  less  negative  with  respect 
to  the  grid  which  Is  operated  at  the  full  neg- 
atlve  vo/foge— 1200  volts.  Focusing  to  the 
desired  sharpness  is  accomplished  by  ad* 
justing  the  confcrcf  on  R23  to  provide  the 
correct  potential  for  anode  no,  I,  Interde* 
pendency  between  the  focus  and  thejnten* 
sity  controls  is  inherent  in  all  electrostatl* 
cally  focused  cathode*ray  tubes.  In  short, 
there  is  an  optimum  setting  of  the  focus  con- 
trol  for  every  setting  of  the  intensity  con- 
trol.  The  second  anode  of  the  5BP1A  is  oper- 
ated at  ground  potential  in  this  instrument. 
Also  one  of  each  pair  of  deflection  plates  is 
operated  at  ground  potential. 

The  cathode  is  operated  at  a high  negative 
potential  (approximately  1200  volts)  so  that 
the  total  overall  accelerating  voltage  of  this 
tube  is  regarded  as  1200  volts  since  the  sec- 
ond onode  Is  operated  at  ground  potential. 
The  vertical  and  horizontal  positioning  con- 
trols which  are  connected  to  their  respective 
deflection  plates  are  capable  of  supplying 
either  a positive  or  negative  d-c  potential 
to  the  deflection  plates.  This  permits  the 
spot  to  be  positioned  at  any  desired  place 
on  the  entire  screen. 


time.  Such  an  arrangement  is  made  possible 
by  the  inclusion  in  the  oscilloscope  of  a Time 
Base^Generator.  Th6  function  of  this  genera- 
tor is  to  move  the  spot  across  the  screen  at  a 
constant  rate  from  left  to  right  between  two 
selected  points,  to  return  die  spot  almost  in- 
stantaneously to  its  original  position,  and  to 


172  The  Oscilloscope 


THE  RADIO 


B 


Figure  4 

SAWTOOTH  WAVE  FORM 


repeat  this  procedure  at  a specified  rate.  This 
action  is  accomplished  by  the  voltage  output 
from  the  time  base  (sweep)  generator.  The 
rate  at  which  this  voltage  repeats  the  cycle 
of  sweeping  the  spot  across  the  screen  is  re- 
ferred to  as  the  sweep  frequency.  The  sweep 
voltage  necessary  to  produce  the  motion  de- 
scribed above  must  be  of  a sawtooth  wave- 
form, such  as  that  shown  in  figure  4. 

The  sweep  occurs  as  the  voltage  varies 
from  A to  B,  and  the  return  trace  as  the  volt- 
age varies  from  B to  C.  If  A-B  is  a straight 
line,  the  sweep  generated  by  this  voltage  will 
be  linear.  It  should  be  realized  that  the  saw- 
tooth sweep  signal  is  only  used  to  plot  varia- 
tions in  the  vertical  axis  signal  with  respect 
to  time.  Specialized  studies  have  made  neces- 
sary the  use  of  sweep  signals  of  various 
shapes  which  are  introduced  from  an  external 
source  through  the  Horizontal  Input  terminals. 

The  Sawtooth  The  sawtooth  voltage  neces- 

Generotor  sary  to  obtain  the  linear  time 

base  is  generated  by  the  cir- 
cuit of  figure  5,  which  operates  as  follows: 

A type  884  gas  ttiode  (Vj)  is  used  for  the 
sweep  generator  tube.  This  tube  contains  an 
inert  gas  which  ionizes  when  the  voltage  be- 
tween the  cathode  and  the  plate  teaches  a cer- 
tain value.  The  ionizing  voltage  depends  upon 
the  bias  voltage  of  the  tube,  which  is  deter- 
mined by  the  voltage  divider  resistors  Rij-Ri,. 
With  a specific  negative  bias  applied  to  the 


884  tube,  the  tube  will  ionize  (or  fire)  at  a 

specific  plate  voltage. 

Capacitors  Cio-C,,  ate  selectively  connected 
in  parallel  with  the  884  tube.  Resistor  R,, 
limits  the  peak  current  drain  of  the  gas  triode. 
The  plate  voltage  on  this  tube  is  obtained 
through  resistors  R^,,  R„  and  R^.  The  voltage 
applied  to  the  plate  of  the  884  tube  cannot  teach 
the  power  supply  voltage  because  of  the  charg- 
ing effect  this  voltage  has  upon  the  capacitor 
which  is  connected  across  the  tube.  This  ca- 
pacitor charges  until  the  plate  voltage  becomes 
high  enough  to  ionize  the  gas  in  the  tube.  At 
this  time,  the  884  tube  starts  to  conduct  and 
the  capacitor  discharges  through  the  tube  until 
its  voltage  falls  to  the  extinction  potential  of 
the  tube.  When  the  tube  stops  conducting,  the 
capacitor  voltage  builds  up  until  the  tube  fires 
again.  As  this  action  continues,  it  results  in 
the  sawtooth  wave  form  of  figure  4 appearing 
at  the  junction  of  R,,  and  Rj,. 

Synchronization  Provision  has  been  made  so 
the  sweep  generator  may  be 
synchronized  from  the  vertical  amplifier  or 
from  an  external  source.  The  switch  Sj  shown 
in  figure  5 is  mounted  on  the  front  panel  to  be 
easily  accessible  to  the  operator. 

If  no  synchronizing  voltage  is  applied,  the 
discharge  tube  will  begin  to  conduct  when  the 
plate  potential  teaches  the  value  of  Ef  (Firing 
Potential).  When  this  breakdown  takes  place 
and  the  tube  begins  to  conduct,  the  capacitor 
is  discharged  rapidly  through  the  tube,  and  the 
plate  voltage  decreases  until  it  reaches  the 
extinction  potential  E^.  At  this  point  conduc- 
tion ceases,  and  the  plate  potential  rises  slow- 
ly as  the  capacitor  begins  to  charge  through 
Rj,  and  Rj,.  The  plate  potential  will  again 
teach  a point  of  conduction  and  the  circuit 
will  start  a new  cycle.  The  rapidity  of  the 
plate  voltage  rise  is  dependent  upon  the  circuit 
constants  Rj,,  Rj,,  and  the  capacitor  selected. 


SCHEMATIC 


Figure  5 
OF  SWEEP 


GENERATOR 


HANDBOOK 


The  Oscilloscope  173 


Eb  + 


C10-C14,  as  well  as  the  supply  voltage  £(,. 
The  exact  relationship  is  given  by: 


Ec=Eb 


Where  E(.==Capacitor  voltage  at  time  t 

Ei,=Supply  voltage  (B+  supply  - cathode 
bias) 

Ef=:Firing  potential  or  potential  at  which 
time-base  gas  ttiode  fires 
Ex=Extinction  potential  or  potential  at 
which  time-base  gas  triode  ceases 
to  conduct 

e=Base  of  natural  logarithms 
t=  Time  in  seconds 
r=Resistance  in  ohms  (Rj,  + Rj,) 
c=Capacity  in  farads  or  „) 


The  frequency  of  oscillation  will  be  approx- 
imately: 


Ef-E, 


Under  this  condition  (no  synchronizing  sig- 
nal applied)  the  oscillator  is  said  to  be  free 
running. 

When  a positive  synchronizing  voltage  is 
applied  to  the  grid,  the  firing  potential  of  the 
tube  is  reduced.  The  tube  therefore  ionizes  at 
a lower  plate  potential  than  when  no  grid  sig- 
nal is  applied.  Thus  the  applied  snychroniz- 
ing  voltage  fires  the  gas-filled  triode  each 


time  the  plate  potential  rises  to  a sufficient 
value,  so  that  the  sweep  recurs  at  the  same 
or  an  integral  sub-multiple  of  the  synchroniz- 
ing signal  rate.  This  is  illustrated  in  figure  6. 


Power  Supply  Figure  7 shows  the  power  sup- 
ply to  be  made  up  of  two  defi- 
nite sections:  a low  voltage  positive  supply 
which  provides  power  for  operating  the  ampli- 
fiers, the  sweep  generator,  and  the  positioning 
circuits  of  the  cathode-ray  tube;  and  the  high 
voltage  negative  supply  which  provides  the 
potentials  necessary  for  operating  the  various 


230  V.  CONN, 

Figure  7 

SCHEMATIC  OF  POWER  SUPPLY 


174  TheOscilloscope  THE  RADIO 


HANDBOOK 


Display  of  Waveforms  175 


Figure  9 

PROJECTION  DRAWING  OF  A SINEWAVE 
APPLIED  TO  THE  VERTICAL  AXIS  AND  A 
SAWTOOTH  WAVE  OF  THE  SAME  FRE- 
QUENCY  APPLIED  SIMULTANEOUSLY  ON 
THE  HORIZONTAL  AXIS 


PROJECTION  DRAWING  SHOWING  THE  RE- 
SULTANT PATTERN  WHEN  THE  FRE- 
QUENCYOF  THE  SAWTOOTH  IS  ONE-HALF 
OF  THAT  EMPLOYED  IN  FIGURE  9 


electrodes  of  the  cathode-ray  tube,  arid  for 
certain  positioning  controls. 

The  positive  low  voltage  supply  consists 
of  full-wave  rectifier  (V,),  the  output  of  which 
is  filtered  by  a capacitor  input  filter  (20-20  /tfd. 
and  8 H).  It  furnishes  approximately  400  volts. 
The  high  voltage  power  supply  employs  a half 
wave  rectifier  tube,  V,.  The  output  of  this  rec- 
tifier is  filtered  by  a resistance-capacitor  fil- 
ter consisting  of  0.5-0. 5 pfd.  and  .18  M.  A 
voltage  divider  network  attached  from  the  out- 
put of  this  filter  obtains  the  proper  operating 
potentials  for  the  various  electrodes  of  the 
cathode-ray  tube.  The  complete  schematic  of 
the  Du  Mont  274-A  Oscilloscope  is  shown  in 
figure  8. 


9-2  Display  of  Waveforms 

Together  with  a working  knowledge  of  the 
controls  of  the  oscilloscope,  an  understanding 
of  how  the  patterns  are  traced  on  the  screen 
must  be  obtained  for  a thorough  knowledge  of 
oscilloscope  operation.  With  this  in  mind  a 
careful  analysis  of  two  fundamental  waveform 
patterns  is  discussed  under  the  following 
headings; 

a.  Patterns  plotted  against  time  (using  the 
sweep  generator  for  horizontal  deflection). 

b.  Lissajous  Figures  (using  a sine  wave  for 
horizontal  deflection). 

Patterns  Plotted  A sine  wave  is  typical  of 
Against  Time  such  a pattern  and  is  con- 

venient for  this  study.  This 


wave  is  amplified  by  the  vertical  amplifier 
and  impressed  on  the  vertical  (Y-axis)  deflec- 
tion plates  of  the  cathode-ray  tube.  Simultane- 
ously the  sawtooth  wave  from  the  time  base 
generator  is  amplified  and  impressed  on  the 
horizontal  (X-axis)  deflection  plates. 

The  electron  beam  moves  in  accordance 
with  the  resultant  of  the  sine  and  sawtooth 
signals.  The  effect  is  shown  in  figure  9 where 
the  sine  and  sawtooth  waves  are  graphically 
represented  on  time  and  voltage  axes.  Points 
on  the  two  waves  that  occur  simultaneously 
are  numbered  similarly.  For  example,  point  2 
on  the  sine  wave  and  point  2 on  the  sawtooth 
wave  occur  at  the  same  instant.  Therefore  the 
position  of  the  beam  at  instant  2 is  the  result- 
ant of  the  voltages  on  the  horizontal  and  ver- 
tical deflection  plates  at  instant  2.  Referring 
to  figure  9,  by  projecting  lines  from  the  two 
point  2 positions,  the  position  of  the  electron 
beam  at  instant  2 can  be  located.  If  projec- 
tions were  drawn  from  every  other  instantane- 
ous position  of  each  wave  to  intersect  on  the 
Oircle  representing  the  tube  screen,  the  inter- 
sections of  similarly  timed  projections  would 
trace  out  a sine  wave. 

In  summation,  figure  9 illustrates  the  prin- 
ciples involved  in  producing  a sine  wave  trace 
on  the  screen  of  a cathode-ray  tube.  Each  in- 
tersection of  similarly  timed  projections  rep- 
resents the  position  of  the  electron  beam  act- 
ing under  the  influence  of  the  varying  voltage 
waveforms  on  each  pair  of  deflection  plates. 
Figure  10  shows  the  effect  on  the  pattern  of 
decreasing  the  frequency  of  the  sawtooth 


176  The  Oscilloscope 


THE  RADIO 


PROJECTION  DRAWING  SHOWING  THE  RE- 
SULTANT LISSAJOUS  PATTERN  WHEN  A 
SINE  WAVE  APPLIED  TO  THE  HORIZON- 
TAL AXIS  IS  THREE  TIMES  THAT  AP- 
PLIED TO  THE  VERTICAL  AXIS 


wave.  Any  recurrent  waveform  plotted  against 
time  can  be  displayed  and  analyzed  by  the 
same  procedure  as  used  in  these  examples. 

The  sine  wave  problem  just  illustrated  is 
typical  of  the  method  by  which  any  waveform 
can  be  displayed  on  the  screen  of  the  cathode- 
ray  tube.  Such  waveforms  as  square  wave, 
sawtooth  wave,  and  many  mote  irregular  recur- 
rent waveforms  can  be  observed  by  the  same 
method  explained  in  the  preceding  paragraphs. 


Lissajous  Figures 

Another  fundamental  pattern  is  the  Lissajous 
figure,  named  after  the  19th  century  French 
scientist.  This  type  of  pattern  is  of  particular 
use  in  determining  the  frequency  ratio  between 
two  sine  wave  signals.  If  one  of  these  signals 
is  known,  the  other  can  be  easily  calculated 
from  the  pattern  made  by  the  two  signals  upon 
the  screen  of  the  cathode-ray  tube.  Common 
practice  is  to  connect  the  known  signal  to  the 
horizontal  channel  and  the  unknown  signal  to 
the  vertical  channel. 

The  presentation  of  Lissajous  figures  can 
be  analyzed  by  the  same  method  as  previously 
used  for  sine  wave  presentation.  A simple  ex- 
ample is  shown  in  figure  11.  The  frequency 
ratio  of  the  signal  on  the  horizontal  axis  to  the 
signal  on  the  vertical  axis  is  3 to  1.  If  the 
known  signal  on  the  horizontal  axis  is  60  cy- 
cles per  second,  the  signal  on  the  vertical 
axis  is  20  cycles. 


Figure  12 

METHOD  OF  CALCULATING  FREQUENCY 
RATIO  OF  LISSAJOUS  FIGURES 


Obtaining  a Lissajous  1.  The  horizontal  am- 
Pattern  on  the  screen  plifiet  should  be  discon- 
Oscilloscope  Settings  nected  from  the  sweep 
oscillator.  The  signal 
to  be  examined  should  be  connected  to  the 
horizontal  amplifier  of  the  oscilloscope. 

2.  An  audio  oscillator  signal  should  be  con- 
nected to  the  vertical  amplifier  of  the  oscillo- 
scope. 

3.  By  adjusting  the  frequency  of  the  audio 
oscillator  a stationary  pattern  should  be  ob- 
tained on  the  screen  of  the  oscilloscope.  It  is 
not  necessary  to  stop  the  pattern,  but  merely 
to  slow  it  up  enough  to  count  the  loops  at  the 
side  of  the  pattern. 

4.  Count  the  number  of  loops  which  intersect 
an  imaginary  vertical  line  AB  and  the  number 
of  loops  which  intersect  the  imaginary  hori- 
zontal line  BC  as  in  figure  12.  The  ratio  of 
the  number  of  loops  which  intersect  AB  is  to 


RATIO  b i 


Figure  13 

OTHER  LISSAJOUS  PATTERNS 


HANDBOOK 


Lissajous  Figures  177 


LISSAJOUS  PATTERNS  OBTAINED  FROM  THE  MAJOR  PHASE  DIFFERENCE  ANGLES 


the  number  of  loops  which  intersect  BC  as  the 
frequency  of  the  horizontal  signal  is  to  the 
frequency  of  the  vertical  signal. 

Figure  13  shows  other  examples  of  Lissa- 
jous figures.  In  each  case  the  frequency  ratio 
shown  is  the  frequency  ratio  of  the  signal  on 
the  horizontal  axis  to  that  on  the  vertical 
axis. 

Phase  Differ-  Coming  under  the  heading  of 
ence  Patterns  Lissajous  figures  is  the  method 
used  to  determine  the  phase 
difference  between  signals  of  the  same  fre- 
quency. The  patterns  involved  take  on  the 
form  of  ellipses  with  different  degrees  of  ec- 
centricity. 

The  following  steps  should  be  taken  to  ob- 
tain a phase-difference  pattern: 

1.  With  no  signal  input  to  the  oscilloscope, 
the  spot  should  be  centered  on  the  screen 
of  the  tube. 

2.  Connect  one  signal  to  the  vertical  ampli- 
fier of  the  oscilloscope,  and  the  other 
signal  to  the  horizontal  amplifier. 

3.  Connect  a common  ground  between  the 
two  frequencies  under  investigation  and 
the  oscilloscope. 

4.  Adjust  the  vertical  amplifier  gain  so  as 
to  give  about  3 inches  of  deflection  on  a 
5 inch  tube,  and  adjust  the  calibrated 
scale  of  the  oscilloscope  so  that  the  ver- 
tical axis  of  the  scale  coincides  precise- 
ly with  the  vertical  deflection  of  the  spot. 

5.  Remove  the  signal  from  the  vertical  am- 
plifier, being  careful  not  to  change  the 
setting  of  the  vertical  gain  control. 

6.  Increase  the  gain  of  the  horizontal  am- 
plifier to  give  a deflection  exactly  the 
same  as  that  to  which  the  vertical  am- 


plifier control  is  adjusted  (3  inches).  Re- 
connect the  signal  to  the  vertical  ampli- 
fier. 

The  resulting  pattern  will  give  an  accurate 
picture  of  the  exact  phase  difference  between 
the  two  waves.  If  these  two  patterns  are  ex- 
actly the  same  frequency  but  different  in  phase 
and  maintain  that  difference,  the  pattern  on 
the  screen  will  remain  stationary.  If,  'however, 
one  of  these  frequencies  is  drifting  slightly, 
the  pattern  will  drift  slowly  through  360®.  The 
phase  angles  of  0°,  45°,  90°,  135°,  180°, 
225°,  270°,  315°  are  shown  in  figure  14. 

Each  of  the  eight  patterns  in  figure  14  can 
be  analyzed  separately  by  the  previously  used 


Figure  15 

PROJECTION  DRAWING  SHOWING  THE  RE- 
SULTANT PHASE  DIFFERENCE  PATTERN 
OF  TWO  SINE  WAVES  45°  OUT  OF  PHASE 


projection  method.  Figure  15  shows  two  sine 
waves  which  differ  in  phase  being  projected 
on  to  the  screen  of  the  cathode-ray  tube.  These 
signals  represent  a phase  difference  of  45°. 
It  is  extremely  important:  (1)  that  the  spot 
has  been  centered  on  the  screen  of  the  cathode- 
ray  tube,  (2)  that  both  the  horizontal  and  ver- 
tical amplifiers  have  been  adjusted  to  give 
exactly  the  same  gain,  and  (3)  that  the  cali- 
brated scale  be  originally  set  to  coincide  with 
the  displacement  of  the  signal  along  the  ver- 
tical axis.  If  the  amplifiers  of  the  oscilloscope 
ate  not  used  for  conveying  the  signal  to  the 
deflection  plates  of  the  cathode-ray  tube,  the 
coarse  frequency  switch  should  be  set  to  hori- 
zontal input  direct  and  the  vertical  input 


switch  to  direct  and  the  outputs  of  the  two 
signals  must  be  adjusted  to  result  in  exactly 
the  same  vertical  deflection  as  horizontal  de- 
flection. Once  this  deflection  has  been  set  by 
either  the  oscillator  output  controls  or  the  am- 
plifier gain  controls  in  the  oscillograph,  it 
should  not  be  changed  for  the  duration  of  the 
measurement. 

Determination  of  The  relation  commonly  used 
the  Phase  Angle  in  determining  the  phase 
angle  between  signals  is: 

Y intercept 

Sine  8 =*=  

Y maximum 


Figure  18 

MODULATED  CARRIER  WAVE  PATTERN 


Figure  19 

PROJECTION  DRAWING  SHOWING  TRAPE- 
ZOIDAL PATTERN 


HANDBOOK 


Trapezoidal  Pattern  179 


Figure  20 

PROJECTION  DRAWING  SHOWING  MODU- 
LATED CARRIER  WAVE  PATTERN 


where  6 = phase  angle  between  signals 

Y intercept  = point  where  ellipse  crosses  ver- 

tical axis  measured  in  tenths  of 
inches.  (Calibrations  on  the 
calibrated  screen) 

Y maximum  = highest  vertical  point  on  ellipse 

in  tenths  of  inches 

Several  examples  of  the  use  of  the  formula  are 
given  in  figure  16.  In  each  case  the  Y inter- 
cept and  Y maximum  are  indicated  together 
with  the  sine  of  the  angle  and  the  angle  itself. 
For  the  operator  to  observe  these  various  pat- 
terns with  a single  signal  source  such  as  the 
test  signal,  there  are  many  types  of  phase 
shifters  which  can  be  used.  Circuits  can  be 
obtained  from  a number  of  radio  text  books. 
The  procedure  is  to  connect  the  original  sig- 
nal to  the  horizontal  channel  of  the  oscillo- 
scope and  the  signal  which  has  passed  through 
the  phase  shifter  to  the  vertical  channel  of 
the  oscilloscope,  and  follow  the  procedure  set 
forth  in  this  discussion  to  observe  the  various 
phase  shift  patterns. 


9-4  Monitoring  Transmitter 

Performance  with  the  Oscilloscope 


The  oscilloscope  may  be  used  as  an  aid  for 
the  proper  operation  of  a radiotelephone  trans- 
mitter, and  may  be  used  as  an  indicator  of  the 
overall  performance  of  the  transmitter  output 
signal,  and  as  a modulation  monitor. 

Waveforms  There  are  two  types  of  patterns 
that  can  serve  as  indicators,  the 
trapezoidal  pattern  (figure  17)  and  the  modu- 


R.F. POWER  AMPLIFIER 


AT  THS  OSCILLOSCOPE  AS  SHOWN. 


Figure  21 

MONITORING  CIRCUIT  FOR  TRAPEZOI- 
DAL MODULATION  PATTERN 


lated  wave  pattern  (figure  18).  The  trapezoidal 
pattern  is  presented  on  the  screen  by  impress- 
ing a modulated  carrier  wave  signal  on  the  ver- 
tical deflection  plates  and  the  signal  that 
modulates  the  carrier  wave  signal  (the  modu- 
lating signal)  on  the  horizontal  deflection 
plates.  The  trapezoidal  pattern  can  be  ana- 
lyzed by  the  method  used  previously  in  analy- 
zing waveforms.  Figure  19  shows  how  the  sig- 
nals cause  the  electron  beam  to  trace  out  the 
pattern. 

The  modulated  wave  pattern  is  accomplished 
by  presenting  a modulated  carrier  wave  on  the 
vertical  deflection  plates  and  by  using  the 
time-base  generator  for  horizontal  deflection. 
The  modulated  wave  pattern  also  can  be  used 
for  analyzing  waveforms.  Figure  20  shows  how 
the  two  signals  cause  the  electron  beam  to 
trace  out  the  pattern. 

The  Trapezoidal  The  oscilloscope  connec- 

Potfern  tions  for  obtaining  a trape- 

zoidal pattern  are  shown  in 
figure  21.  A portion  of  the  audio  output  of  the 
transmitter  modulator  is  applied  to  the  hori- 
zontal input  of  the  oscilloscope.  The  vertical 
amplifier  of  the  oscilloscope  is  disconnected, 
and  a small  amount  of  modulated  r-f  energy  is 
coupled  directly  to  the  vertical  deflection 
plates  of  the  oscilloscope.  A small  pickup 

loop,  loosely  coupled  to  the  final  amplifier 
tank  circuit  and  connected  to  the  vertical  de- 


180  The  Oscilloscope 


THE  RADIO 


TRAPEZOIDAL  WAVE  PATTERN 


Figure  22 


Figure  23 


Figure  24 


(LESS  THAN  100%  MODULATION)  (100%  MODULATION)  (OVER  MODULATION) 


flection  plates  by  a short  length  of  coaxial 
line  will  suffice.  The  amount  of  excitation  to 
the  plates  of  the  oscilloscope  may  be  adjusted 
to  provide  a pattern  of  convenient  size.  Upon 
modulation  of  the  transmittet,  the  trapezoidal 
pattern  will  appear.  By  changing  the  degree  of 
modulation  of  the  carrier  wave  the  shape  of 
the  pattern  will  change.  Figures  22  and  23 
show  the  trapezoidal  pattern  for  various  de- 
grees of  modulation.  The  percentage  of  modu- 
lation may  be  determined  by  the  following  for- 
mula; 


Modulation  percentage 
Tmax  ~ Emin 
Tmax  1"  Tm}n 


X 100 


where  and  Ef„in  are  defined  as  in 

figure  22. 

An  overmodulated  signal  is  shown  in  figure 


24. 


The  Modulated  The  oscilloscope  connections 

Wove  Pattern  for  obtaining  a modulated 

wave  pattern  are  shown  in 


R.F.  POWER  AMPLIFIER 


CRO 


BF  FROM 
MODULATOR 


LC  TUNES  TO  OP- 
ERATING FREQUENCY 


Figure  25 

MONITORING  CIRCUIT  FOR 
MODULATED  WAVE  PATTERN 


figure  25.  The  internal  sweep  circuit  of  the 
oscilloscope  is  applied  to  the  horizontal 
plates,  and  the  modulated  t-f  signal  is  applied 
to  the  vertical  plates,  as  described  before.  If 
desired,  the  internal  sweep  circuit  may  be  sny- 
chronized  with  the  modulating  signal  of  the 
transmittet  by  applying  a small  portion  of  the 
modulator  output  signal  to  the  external  sync 
post  of  the  oscilloscope.  The  percentage  of 
modulation  may  be  determined  in  the  same 
fashion  as  with  a trapezoidal  pattern.  Figures 
26,  27  and  28  show  the  modulated  wave  pat- 
tern for  various  degrees  of  modulation. 


9-5  Receiver  l-F  Alignment 
with  an  Oscilloscope 


The  alignment  of  the  i-f  amplifiers  of  a re- 
ceiver consists  of  adjusting  all  the  tuned  cir- 
cuits to  resonance  at  the  intermediate  frequen- 
cy and  at  the  same  time  to  permit  passage  of 
a predetermined  number  of  side  bands.  The 
best  indication  of  this  adjustment  is  a reso- 
nance curve  representing  the  response  of  the 
i-f  circuit  to  its  particular  range  of  frequencies. 

As  a rule  medium  and  low-priced  receivers 
use  i-f  transformers  whose  bandwidth  is  about 
5 kc.  on  each  side  of  the  fundamental  frequen- 
cy. The  response  curve  of  these  i-f  transform- 
ers is  shown  in  figure  29.  High  fidelity  re- 
ceivers usually  contain  i-f  transformers  which 
have  a broader  bandwidth  which  is  usually  10 
kc.  on  each  side  of  the  fundamental.  The  re- 
sponse curve  for  this  type  transformer  is  shown 
in  figure  30. 

Resonance  curves  such  as  these  can  be  dis- 
played on  the  screen  of  an  oscilloscope.  For 
a complete  understanding  of  the  procedure  it 
is  important  to  know  how  the  resonance  curve 
is  traced. 


HANDBOOK 


Receiver  Alignment  181 


CARRIER  WAVE  PATTERN 


Figure  26 


Figure  27 


Figure  28 


(LESS  THAN  100%  MODULATION)  O00%  MODULATION) 


(OVER  MODULATION) 


The  Resonance  To  present  a resonance  curve 
Curve  on  the  on  the  screen,  a frequency- 
Screen  modulated  signal  source  must 

be  available.  This  signal 
source  is  a signal  generator  whose  output  is 
the  fundamental  i-f  frequency  which  is  fre- 
quency-modulated 5 to  10  kc.  each  side  of  the 
fundamental  frequency.  A signal  generator  of 
this  type  generally  takes  the  form  of  an  ordi- 
nary signal  generator  with  a rotating  motor 
driven  tuned  circuit  capacitor,  called  a wob- 


Figure  29 

FREQUENCY  RESPONSE  CURVE  OF  THE 
I-F  OF  A LOW  PRICED  RECEIVER 


Figure  30 

FREQUENCY  RESPONSE  OF 
HIGH-FIDELITY  I-F  SYSTEM 


bulator,  or  its  electronic  equivalent,  a react- 
ance tube. 

The  method  of  presenting  a resonance  curve 
on  the  screen  is  to  connect  the  vertical  chan- 
nel of  the  oscilloscope  across  the  detector 
load  of  the  receiver  as  shown  in  the  detectors 
of  figure  31  (between  point  A and  ground)  and 
the  time-base  generator  output  to  the  horizon- 
tal channel.  In  this  way  the  d-c  voltage  across 
the  detector  load  varies  with  the  frequencies 
which  are  passed  by  the  i-f  system.  Thus,  if 
the  time-base  generator  is  set  at  the  frequency 
of  rotation  of  the  motor  driven  capacitor,  or 
the  reactance  tube,  a pattern  resembling  fig- 
ure 32,  a double  resonance  curve,  appears  on 
the  screen. 

Figure  32  is  explained  by  considering  fig- 
ure 33.  In  half  a rotation  of  the  motor  driven 
capacitor  the  frequency  increases  from  445 
kc.  to  465  kc.,  more  than  covering  the  range 
of  frequencies  passed  by  the  i-f  system. 
Therefore,  a full  resonance  curve  is  presented 
on  the  screen  during  this  half  cycle  of  rota- 
tion since  only  half  a cycle  of  the  voltage  pro- 
ducing horizontal  deflection  has  transpired. 
In  the  second  half  of  the  rotation  the  motor 


Figure  31 

CONNECTION  OF  THE  OSCILLOSCOPE 
ACROSS  THE  DETECTOR  LOAD 


182  The  Oscilloscope 


THE  RADIO 


Figure  32 

DOUBLE  RESONANCE  CURVE 


Figure  33 

DOUBLE  RESONANCE  ACHIEVED  BY 
COMPLETE  ROTATION  OF  THE  MOTOR 
DRIVEN  CAPACITOR 


Figure  34 

SUPER-POSITION  OF  RESONANCE  CURVES 


driven  capacitor  takes  the  frequency  of  the 
signal  in  the  reverse  order  through  the  range 
of  frequencies  passed  by  the  i-f  system.  In 
this  interval  the  time-base  generator  sawtooth 
waveform  completes  its  cycle,  dtawing  the 
electron  beam  further  across  the  screen  and 
then  returning  it  to  the  starting  point.  Subse- 
quent cycles  of  the  motor  driven  capacitor  and 
the  sawtooth  voltage  merely  retrace  the  same 
pattern.  Since  the  signal  being  viewed  is  ap- 
plied through  the  vertical  amplifier,  the  sweep 
can  be  synchronized  internally. 

Some  signal  generators,  particularly  those 
employing  a reactance  tube,  provide  a sweep 
output  in  the  form  of  a sine  wave  which  is 
synchronized  to  the  frequency  with  which  the 
reactance  tube  is  swinging  the  fundamental 
frequency  through  its  limits,  usually  60  cycles 
per  second.  If  such  a signal  is  used  for  hori- 
zontal deflection,  it  is  already  synchronized. 
Since  this  signal  is  a sine  wave,  the  response 


curve  is  observed  as  it  sweeps  the  spot  across 
the  screen  from  left  to  right;  and  it  is  observed 
again  as  the  sine  wave  sweeps  the  spot  back 
again  from  right  to  left.  Under  these  condi- 
tions the  two  response  curves  are  superim- 
posed on  each  other  and  the  high  frequency 
responses  of  both  curves  are  at  one  end  and 
the  low  frequency  response  of  both  curves  is 
at  the  other  end.  The  i-f  trimmer  capacitors 
are  adjusted  to  produce  a response  curve 
which  is  symmetrical  on  each  side  of  the  fun- 
damental frequency. 

When  using  sawtooth  sweep,  the  two  re- 
sponse curves  can  also  be  superimposed.  If 
the  sawtooth  signal  is  generated  at  exactly 
twice  the  frequency  of  rotation  of  the  motor 
driven  capacitor,  the  two  resonance  curves 
will  be  superimposed  (figure  34)  if  the  i-f 
transformers  are  properly  tuned.  If  the  two 
curves  do  not  coincide  the  i-f  trimmer  capaci- 
tors should  be  adjusted.  At  the  point  of  co- 
incidence the  tuning  is  correct.  It  should  be 
pointed  out  that  rarely  do  the  two  curves  agree 
perfectly.  As  a result,  optimum  adjustment  is 
made  by  making  the  peaks  coincide.  This  lat- 
ter procedure  is  the  one  generally  used  in  i-f 
adjustment.  When  the  two  curves  coincide,  it 
is  evident  that  the  i-f  system  responds  equal- 
ly to  signals  higher  and  lower  than  the  funda- 
mental i-f  frequency. 

9-6  Single  Sideband  Applications 

Measurement  of  power  output  and  distortion 
are  of  particular  importance  in  SSB  transmitter 
adjustment.  These  measurements  are  related  to 
the  extent  that  distortion  rises  rapidly  when 
the  power  amplifier  is  overloaded.  The  useable 
power  output  of  a SSB  transmitter  is  often  de- 
fined as  the  maximum  peak  envelope  power 


Figure  35 

SINGLE  TONE  PRESENTATION 

Oscilloscope  trace  of  SSB  signal 
modulated  by  single  tone  (A). 
Incomplete  carrier  supression  or 
spurious  products  will  show 
modulated  envelope  of  (B).  The 
ratio  of  supression  is.* 

S = 20  log  A+B 
A-B 


HANDBOOK 


S.S.B.  Applications  183 


OSCILLOSCOPE 


Figure  36 

BLOCK  DIAGRAM  OF 
LINEARITY  TRACER 


obtainable  with  a specified  signal-to-distortion 
ratio.  The  oscilloscope  is  a useful  instrument 
for  measuring  and  studying  distortion  of  all 
types  that  may  be  generated  in  single  sideband 
equipment. 

Single  Tone  When  a SSB  transmitter  is  modu- 
Observations  lated  with  a single  audio  tone, 
the  r-f  output  should  be  a single 
radio  frequency.  If  the  vertical  plates  of  the 
oscilloscope  are  coupled  to  the  output  of  the 
transmitter,  and  the  horizontal  amplifier  sweep 
is  set  to  a slow  rate,  the  scope  presentation 
will  be  as  shown  in  figure  35*  If  unwanted  dis- 
tortion products  or  carrier  are  present,  the  top 
and  bottom  of  the  pattern  will  develop  a **rip- 
ple*’  proportional  to  the  degree  of  spurious 
products. 


The  Linearity  The  linearity  tracer  is  an  aux- 
Tracer  iliary  detector  to  be  used  with 

an  oscilloscope  for  quick  ob- 
servation of  amplifier  adjustments  and  para- 
meter variations.  This  instrument  consists  of 
two  SSB  envelope  detectors  the  outputs  of 
which  connect  to  the  horizontal  and  vertical 
inputs  of  an  oscilloscope.  Figure  36  shows  a 
block  diagram  of  atypical  linearity  test  set-up. 
A two-tone  test  signal  is  normally  employed 
to  supply  a SSB  modulation  envelope,  but  any 
modulating  signal  that  provides  an  envelope 
that  varies  from  zero  to  full  amplitude  may  be 
used.  Speech  modulation  gives  a satisfactory 
trace,  so  that  this  instrument  may  be  used  as 
a visual  monitor  of  transmitter  linearity.  It  is 
particularly  useful  for  monitoring  the  signal 
level  and  clearly  shows  when  the  amplifier 
under  observation  is  overloaded.  The  linearity 
trace  will  be  a straight  line  regardless  of  the 
envelope  shape  if  the  amplifier  has  no  dis- 
tortion. Overloading  causes  a sharp  break  in 
the  linearity  curve.  Distortion  due  to  too  much 
bias  is  also  easily  observed  and  the  adjustment 
for  low  distortion  can  easily  be  made. 

Another  feature  of  the  linearity  detector  is 


R-F  sse  INPUT 
Ft-OM  VOLTAGE 
DIVIDER  OR 
PICKUPCOIL 


GERMANIUM  E.5MM 
DIODE  RFC 


mhL  -L 

'^01  47^ 


AUDIO  OUTPUT 
TO  OSCILLOSCOPE 


Figure  37 
SCHEMATIC  OF 
ENVELOPE  DETECTOR 


that  the  distortion  of  each  individual  stage 
can  be  observed.  This  is  helpful  in  trouble- 
shooting. By  connecting  the  input  envelope 
detector  to  the  output  of  the  SSB  generator, 
the  overall  distortion  of  the  entire  r-f  circuit 
beyond  this  point  is  observed.  The  unit  can 
also  serve  as  a voltage  indicator  which  is 
useful  in  making  tuning  adjustments. 

The  circuit  of  a typical  envelope  detector 
is  shown  in  figure  37.  Two  matched  germainum 
diodes  are  used  as  detectors.  The  detectors 
are  not  linear  at  low  signal  levels,  but  if  the 
nonlinearity  of  the  two  detectors  is  matched, 
the  effect  of  their  nonlinearity  on  the  oscillo- 
scope trace  is  cancelled.  The  effect  of  diode 
differences  is  minimized  by  using  a diode  load 
of  5,000  to  10,000  ohms,  as  shown.  It  is  im- 
portant that  both  detectors  operate  at  approxi- 
mately the  same  signal  level  so  that  their 
differences  will  cancel  more  exactly.  The 
operating  level  should  be  1-volt  or  higher. 

It  is  convenient  to  build  the  detector  in  a 
small  shielded  enclosure  such  as  an  i-f  trans- 
former can  fitted  with  coaxial  input  and  output 
connectors.  Voltage  dividers  can  be  similarly 
constructed  so  that  it  is  easy  to  insert  the  de- 
sired amount  of  voltage  attenuation  from  the 
various  sources.  In  some  cases  it  is  convenient 
to  use  a pickup  loop  on  the  end  of  a short 
length  of  coaxial  cable. 

The  phase  shift  of  the  amplifiers  in  the  os- 
cilloscope should  be  the  same  and  their  fre- 
quency response  should  be  flat  out  to  at  least 
twenty  times  the  frequency  difference  of  the 
two  test  tones.  Excellent  high  frequency  charac- 
teristics are  necessary  because  the  rectified 
SSB  envelope  contains  harmonics  extending 
tothe  limit  of  the  envelope  detector’s  response. 
Inadequate  frequency  response  of  the  vertical 
amplifier  may  cause  a little  "foot”  to  appear 
on  the  lower  end  of  the  trace,  as  shown  in 
figure  38.  If  it  is  small,  it  may  be  safely  neg- 
lected. 

Another  spurious  effect  often  encountered 
is  a double  trace,  as  shown  in  figure  39-  This 
can  usually  be  corrected  with  an  R-C  network 
placed  between  one  detector  and  the  oscillo- 
scope. The  best  method  of  testing  the  detectors 
and  the  amplifiers  is  to  connect  the  input  of 


184  The  Oscilloscope 


Figure  38 

EFFECT  OF  INADEQUATE 
RESPONSE  OF  VERTICAL 
AMPLIFIER 


OUTPUT 

SIGNAL 

LEVEL 


Figure  41 

ORDINATES  ON  LINEARITY 
CURVE  FOR  3RD  ORDER 
DISTORTION  EQUATION 


Figure  39 
DOUBLE  TRACE 
CAUSED  BY  PHASE 
SHIFT 


the  envelope  detectors  in  parallel.  A perfectly 
straight  line  trace  will  result  when  everything 
is  working  properly.  One  detector  is  then  con- 
nected to  the  other  r-f  source  through  a voltage 
divider  adjusted  so  that  no  appreciable  change 
in  the  setting  of  the  oscilloscope  amplifier 
controls  is  required.  Figure  40  illustrates  some 
typical  linearity  traces.  Trace  A is  caused  by 
inadequate  static  plate  current  in  class  A or 
class  B amplifiers  or  a mixer  stage.  To  regain 
linearity,  the  grid  bias  of  the  stage  should  be 
reduced,  the  screen  voltage  should  be  raised, 
or  the  signal  level  should  be  decreased.  Trace 
B is  a result  of  poor  grid  circuit  regulation 
when  grid  current  is  drawn,  or  a result  of  non- 


linear plate  characteristics  of  the  amplifier 
tube  at  large  plate  swings.  Mote  grid  swamping 
should  be  used,  or  the  exciting  signal  should 
be  reduced.  A combination  of  the  effects  of  A 
and  B ate  shown  in  Trace  C.  Trace  D illustrates 
amplifier  overloading.  The  exciting  signal 
should  be  reduced. 

A means  of  estimating  the  distortion  level 
observed  is  quite  useful.  The  first  and  third 
order  distortion  components  may  be  derived  by 
an  equation  that  will  give  the  approximate 
signal-to-distortion  level  ratio  of  a two  tone 
test  signal,  operating  on  a given  linearity  curve. 
Figure  41  shows  a linearity  curve  with  two 
ordinates  erected  at  haH  and  full  peak  input 
signal  level.  The  length  of  the  ordinates  ei 
and  e2  may  be  scaled  and  used  in  the  following 
equation: 

Signal-to-distortion  ratio  in  db  = 20  log  8 ei-e2 

2 ei-e2 


TYPICAL  LINEARITY  TRACES 


Figure  40 

TYPICAL  LINEARITY 
TRACES 


CHAPTER  TEN 


Special  Vacuum  Tube  Circuits 


A whole  new  concept  of  vacuum  tube  appli- 
cations has  been  developed  in  recent  years. 
No  longer  are  vacuum  tubes  chained  to  the 
field  of  communication.  This  chapter  is  de- 
voted to  some  of  the  more  common  circuits  en- 
countered in  industritd  and  military  applica- 
tions of  the  vacuum  tube. 


10-1  Limiting  Circuits 

The  term  limiting  refers  to  the  removal  or 
suppression  by  electronic  means  of  the  ex- 
tremities of  an  electronic  signal.  Circuits 
which  perform  this  function  are  referred  to  as 
limiters  or  clippers.  Limiters  are  useful  in 
wave-shaping  circuits  where  it  is  desirable  to 
square  off  the  extremities  of  the  applied  sig- 
nal. A sine  wave  may  be  applied  to  a limiter 
circuit  to  produce  a rectangular  wave.  A 
peaked  wave  may  be  applied  to  a limiter  cir- 
cuit to  eliminate  either  the  positive  or  nega- 
tive peaks  from  the  output.  Limiter  circuits 
are  employed  in  FM  receivers  where  it  is  nec- 
essary to  limit  the  amplitude  of  the  signal  ap- 
plied to  the  detector.  Limiters  may  be  used  to 
reduce  automobile  ignition  noise  in  short-wave 
receivers,  or  to  maintain  a high  average  level 
of  modulation  in  a transmitter.  They  may  also 
be  used  as  protective  devices  to  limit  input 
signals  to  special  circuits. 


Diode  Limiters  The  characteristics  of  a 
diode  tube  are  such  that  the 
tube  conducts  only  when  the  plate  is  at  a posi- 
tive potential  with  respect  to  the  cathode.  A 
positive  potential  may  be  placed  on  the  cath- 
ode, but  the  tube  will  not  conduct  until  the 
voltage  on  the  plate  rises  above  an  equally 
positive  value.  As  the  plate  becomes  more 
positive  with  respect  to  the  cathode,  the  diode 
conducts  and  passes  that  portion  of  the  wave 
that  is  more  positive  than  the  cathode  voltage. 
Diodes  may  be  used  as  either  series  or  paral- 
lel limiters,  as  shown  in  figure  1.  A diode  may 
be  so  biased  that  only  a certain  portion  of  the 
positive  or  negative  cycle  is  removed. 

Audio  Peak  An  audio  peak  clipper  consisting 
Limiting  of  two  diode  limiters  maybe  used 

to  limit  the  amplitude  of  an  au- 
dio signal  to  a predetermined  value  to  provide 
a high  average  level  of  modulation  without 
danger  of  overmodulation.  An  effective  limiter 
for  this  service  is  the  series-diode  gate  clip- 
per. A circuit  of  this  clipper  is  shown  in  fig- 
ure 2.  The  audio  signal  to  be  clipped  is  cou- 
pled to  the  clipper  through  C,.  R,  and  are 
the  clipper  input  and  output  load  resistors. 
The  clipper  plates  are  tied  together  and  are 
connected  to  the  clipping  level  control,  R„ 
through  the  series  resistor,  R,.  R4  acts  as  a 
voltage  divider  between  the  high  voltage  sup- 
ply and  ground.  The  exact  point  at  which  clip- 


185 


186  Special  Vacuum  Tube  Circuits 


THE  RADIO 


Figure  1 

VARIOUS  DIODE  LIMITING  CIRCUITS 

Seri  es  diodes  limiting  positive  and  negative  peaks  are  shown  in  A and  B.  Parallel  diodes  limit- 
ing positive  and  negative  peaks  are  shown  in  C and  0.  Parallel  diodes  limiting  above  and  below 
ground  are  shown  in  E and  F,  Parallel  diode  limiters  which  pass  negative  and  positive  peaks 

are  shown  in  C and  H. 


ping  will  occur  is  set  by  R,,  which  controls  the 
positive  potential  applied  to  the  diode  plates. 

Under  static  conditions,  a d-c  voltage  is  ob- 
tained from  R4  and  applied  through  R,  to  both 
plates  of  the  6AL5  tube.  Current  flows  through 
R4,  R3,  and  divides  through  the  two  diode 
sections  of  the  6AL5  and  the  two  load  resis- 
tors, R,  and  Rj.  All  parts  of  the  clipper  circuit 
are  maintained  at  a positive  potential  above 
ground.  The  voltage  drop  between  the  plate 
and  cathode  of  each  diode  is  very  small  com- 
pared to  the  drop  across  the  300,000-ohm  re- 
sistor (R3)  in  series  with  the  diode  plates. 
The  plate  and  cathode  of  each  diode  are  there- 
fore maintained  at  approximately  equal  poten- 
tials as  long  as  there  is  plate  current  flow. 
Clipping  does  not  occur  until  the  peak  audio 
input  voltage  teaches  a value  greater  than  the 
static  voltages  at  the  plates  of  the  diode. 


Assume  that  R,  has  been  set  to  a point  that 
will  give  4 volts  at  the  plates  of  the  6AL5. 
When  the  peak  audio  input  voltage  is  less  than 
4 volts,  both  halves  of  the  tube  conduct  at  all 
times.  As  long  as  the  tube  conducts,  its  re- 
sistance is  very  low  compared  with  the  plate 
resistor  R,.  Whenever  a voltage  change  occurs 
across  input  resistor  Ri,  the  voltage  at  all  of 
the  tube  elements  increases  or  decreases  by 
the  same  amount  as  the  input  voltage  change, 
and  the  voltage  drop  across  R3  changes  by  an 
equal  amount.  As  long  as  the  peak  input  volt- 
age is  less  than  4 volts,  the  6AL5  acts  merely 
as  a conductor,  and  the  output  cathode  is  per- 
mitted to  follow  all  voltage  changes  at  the  in- 
put cathode. 

If,  under  static  conditions,  4 volts  appear  at 
the  diode  plates,  then  twice  this  voltage  (8 
volts)  will  appear  if  one  of  the  diode  circuits 


HANDBOOK 


Clamping  Circuits  187 


AAr 


t 


Figure  2 

THE  SERIES-DIODE  GATE  CLIPPER  FOR 
AUDIO  PEAK  LIMITING 


Figure  3 

GRID  LIMITING  CIRCUIT 


is  opened,  removing  its  d-c  load  from  the  cir- 
cuit. As  long  as  only  one  of  the  diodes  con- 
tinues to  conduct,  the  voltage  at  the  diode 
plates  cannot  rise  above  twice  the  voltage  se- 
lected by  R4.  In  this  example,  the  voltage  can- 
not rise  above  8 volts.  Now,  if  the  input  audio 
voltage  applied  through  Cj  is  increased  to  any 
peak  value  between  zero  and  plus  4 volts,  the 
first  cathode  of  the  6AL5  will  increase  in  volt- 
age by  the  same  amount  to  the  proper  value  be- 
tween 4 and  8 volts.  The  other  tube  elements 
will  assume  the  same  potential  as  the  first 
cathode.  However,  the  6AL5  plates  cannot  in- 
crease more  than  4 volts  above  their  original 
4-volt  static  level.  When  the  input  voltage  to 
the  first  cathode  of  the  6AL5  increases  to 
more  than  plus  4 volts,  the  cathode  potential 
increases  to  more  than  8 volts.  Since  the  plate 
circuit  potential  remains  at  8 volts,  the  first 
diode  section  ceases  to  conduct  until  the  in- 
put voltage  across  Ri  drops  below  4 volts. 

When  the  input  volr.age  swings  in  a negative 
direction,  it  will  subtract  from  the  4-volt  drop 
across  R,  and  decrease  the  voltage  on  the  in- 
put cathode  by  an  amount  equal  to  the  input 
voltage.  The  plates  and  the  output  cathode  will 
follow  the  voltage  level  at  the  input  cathode 
as  long  as  the  input  voltage  does  not  swing 
below  minus  4 volts.  If  the  input  voltage  does 
not  change  more  than  4 volts  in  a negative 
direction,  the  plates  of  the  6AL5  will  also  be- 
come negative.  The  potential  at  the  output 
cathode  will  follow  the  input  cathode  voltage 
and  decrease  from  its  normal  value  of  4 volts 
until  it  teaches  zero  potential.  As  the  input 
cathode  voltage  decreases  to  less  than  zero, 


the  plates  will  follow.  Howe.ver,  the  output 
cathode,  grounded  through  Rj,  will  stop  at  zero 
potential  as  the  plate  becomes  negative.  Con- 
duction through  the  second  diode  is  impossible 
under  these  conditions.  The  output  cathode 
remains  at  zero  potential  until  the  voltage  at 
the  input  cathode  swings  back  to  zero. 

The  voltage  developed  across  output  resis- 
tor Rj  follows  the  input  voltage  variations  as 
long  as  the  input  voltage  does  not  swing  to  a 
peak  value  greater  than  the  static  voltage  at 
the  diode  plates,  determined  by  R4.  Effective 
clipping  may  thus  be  obtained  at  any  desired 
level. 

The  square-topped  audio  waves  generated 
by  this  clipper  are  high  in  harmonic  content, 
but  these  higher  order  harmonics  may  be  great- 
ly reduced  by  a low-level  speech  filter. 

Grid  Limiters  A triode  grid  limiter  is  shown 
in  figure  3.  On  positive  peaks 
of  the  input  signal,  the  triode  grid  attempts  to 
swing  positive,  and  the  grid-cathode  resist- 
ance drops  to  a value  on  the  order  of  1000 
ohms  or  so.  The  voltage  drop  across  R (usu- 
ally of  the  order  of  1 megohm)  is  large  com- 
pared to  the  grid-cathode  drop,  and  the  result- 
ing limiting  action  removes  the  top  part  of  the 
positive  input  wave. 


10-2  Clomping  Circuits 

A circuit  which  holds  either  amplitude  ex- 
treme of  a waveform  to  a given  reference  level 


eiN — 1(- 


°TJUL" 

100-  - I — i-~i — I-  - I — 


DiOOE  CONDUCTS 


(a)  positive  clamping  circuit 


eiN 


eouT 


® NEGATIVE  CLAMPING  CIRCUIT 


Figure  4 

SIMPLE  POSITIVE  AND  NEGATIVE  CLAMPING  CIRCUITS 


188  Special  Vacuum  Tube  Circuits 


THE  RADIO 


eiN It 


Figure  5 

NEGATIVE  CLAMPING  CIRCUIT  EM- 
PLOYED IN  ELECTROMAGNETIC  SWEEP 
SYSTEM 


B+ 


Cz  DISCHARGE  PATH 


Figure  7 

THE  CHARGE  AND  DISCHARGE  PATHS 
IN  FREE-RUNNING  MULTIVIBRATOR  OF 
FIGURE  6 


Bi- 


Figure  6 

BASIC  MULTIVIBRATOR  CIRCUIT 


of  potential  is  called  a clamping  circuit  or  a 
d-c  restorer.  Clamping  circuits  ate  used  after 
RC  cpupling  circuits  where  the  waveform 
swing  is  required  to  be  either  above  or  below 
the  reference  voltage,  instead  of  alternating 
on  both  sides  of  it  (figure  4).  Clamping  cir- 
cuits are  usually  encountered  in  oscilloscope 
sweep  circuits.  If  the  sweep  voltage  does  not 
always  start  from  the  same  reference  point, 
the  trace  on  the  screen  does  not  begin  at  the 
same  point  on  the  screen  each  time  the  sweep 


is  repeated  and  therefore  is  "jittery.”  If  a 
clamping  circuit  is  placed  between  the  sweep 
amplifier  and  the  deflection  element,  the  start 
of  the  sweep  can  be  regulated  by  adjusting  the 
d-c  voltage  applied  to  the  clamping  tube  (fig- 
ure 5). 


10-3  Multivibrators 

The  multivibrator,  or  relaxation  oscillator, 
is  used  for  the  generation  of  nonsinusoidal 
waveforms.  The  output  is  rich  in  harmonics, 
but  the  inherent  frequency  stability  is  poor. 
The  multivibrator  may  be  stabilized  by  the 
introduction  of  synchronizing  voltages  of  har- 
monic or  subharmonic  frequency. 

In  its  simplest  form,  the  multivibrator  is  a 
simple  two-stage  resistance-capacitance  cou- 
pled amplifier  with  the  output  of  the  second 
stage  coupled  through  a capacitor  to  the  grid 
of  the  first  tube,  as  shown  in  figure  6.  Since 
the  output  of  the  second  stage  is  of  the  proper 
polarity  to  reinforce  the  input  signal  applied 
to  the  first  tube,  oscillations  can  readily  take 
place,  started  by  thermal  agitation  noise  and 


B-f 


@ 


DIRECT-COUPLED  CATHODE 
MULTIVIBRATOR 


Figure  8 

VARIOUS  FORMS  OF  MULTIVIBRATOR  CIRCUITS 


HANDBOOK 


Multivibrators  189 


CIRCUIT 


Figure  9 

ECCLES-JORDAN  MULTIVIBRATOR  CIRCUITS 


miscellaneous  tube  noise.  Oscillation  is  main- 
tained by  the  process  of  building  up  and  dis- 
charging the  store  of  energy  in  the  grid  cou- 
pling capacitors  of  the  two  tubes.  The  charg- 
ing and  discharging  paths  are  shown  in  figure 
7.  Various  forms  of  multivibrators  are  shown 
in  figure  8. 

The  output  of  a multivibrator  may  be  used 
as  a source  of  square  waves,  as  an  electronic 
switch,  or  as  a means  of  obtaining  frequency 
division.  Submultiple  frequencies  as  low  as 
one-tenth  of  the  injected  synchronizing  fre- 
quency may  easily  be  obtained. 

The  Eccles- Jordan  The  Eccles- Jordan  trigger 
Circuit  circuit  is  shown  in  figure 

9A.  This  is  not  a true  mul- 
tivibrator, but  rather  a circuit  that  possesses 
two  conditions  of  stable  equilibrium.  One  con- 
dition is  when  V,  is  conducting  and  Vj  is  cut- 
off; the  other  when  Vj  is  conducting  and  V,  is 
cutoff.  The  circuit  remains  in  one  or  the  other 
of  these  two  stable  conditions  with  no  change 
in  operating  potentials  until  some  external 
action  occurs  which  causes  the  nonconducting 
tube  to  conduct.  The  tubes  then  reverse  their 
functions  and  remain  in  the  new  condition  as 
long  as  no  plate  current  flows  in  the  cutoff 
tube.  This  type  of  circuit  is  known  as  a flip- 
flop  circuit. 


CUTOFF 

TIME 


Figure  10 

SINGLE-SWING  BLOCKING  OSCILLATOR 


Figure  9B  illustrates  a modified  Eccles- 
Jordan  circuit  which  accomplishes  a complete 
cycle  when  triggered  with  a positive  pulse. 
Such  a circuit  is  called  a one-shot  multivibra- 
tor. For  initial  action,  V,  is  cutoff  and  Vj  is 
conducting.  A large  positive  pulse  applied  to 
the  grid  of  V,  causes  this  tube  to  conduct,  and 
the  voltage  at  its  plate  decreases  by  virtue  of 
the  IR  drop  through  Rj.  Capacitor  C2  is  charged 
rapidly  by  this  abrupt  change  in  V,  plate  volt- 
age, and  Vj  becomes  cutoff  while  Vj  conducts. 
This  condition  exists  until  C,  discharges,  al- 
lowing Vj  to  conduct,  raising  the  cathode  bias 
of  V,  until  it  is  once  again  cutoff. 

A direct,  cathode-coupled  multivibrator  is 
shown  in  figure  8A.  Rr  is  a common  cathode 
resistor  for  the  two  tubes,  and  coupling  takes 
place  across  this  resistor.  It  is  impossible  for 
a tube  in  this  circuit  to  completely  cutoff  the 
other  tube,  and  a circuit  of  this  type  is  called 
a free-running  multivibrator  in  which  the  con- 
dition of  one  tube  temporarily  cuts  off  the 
other. 


HARTLEY  OSCILLATOR  USED  AS  BLOCKING 
OSCILLATOR  BY  PROPER  CHOICE  OF  Rj-C, 


190  Special  Vacuum  Tube  Circuits 


THE  RADIO 


@ 

POSITIVE  COUNTING  CIRCUIT 


NEGATIVE  COUNTING  CIRCUIT 


© 

POSITIVE  COUNTING  CIRCUIT  WITH 
METER  INDICATION 


Figure  12 

POSITIVE  AND  NEGATIVE  COUNTING  CIRCUITS 


10-4  The  Blocking  Oscillator 

A blocking  oscillator  is  any  oscillator  which 
cuts  itself  off  after  one  or  more  cycles  caused 
by  the  accumulation  of  a negative  charge  on 
the  grid  capacitor.  This  negative  charge  may 
gradually  be  drained  off  through  the  grid  re- 
sistor of  the  tube,  allowing  the  circuit  to  os- 
cillate once  again.  The  process  is  repeated 
and  the  tube  becomes  an  intermittent  oscilla- 
tor. The  rate  of  such  an  occurance  is  deter- 
mined by  the  R-C  time  constant  of  the  grid  cir- 
cuit. A single-swing  blocking  oscillator  is 
shown  in  figure  10,  wherein  the  tube  is  cutoff 
before  the  completion  of  one  cycle.  The  tube 
produces  single  pulses  of  energy,  the  time 
between  the  pulses  being  regulated  by  the 
discharge  time  of  the  grid  R-C  network.  The 
self-pulsing  blocking  oscillator  is  shown  in 
figure  11,  and  is  used  to  produce  pulses 
of  r-f  energy,  the  number  of  pulses  being  de- 
termined by  the  timing  network  in  the  grid  cir- 
cuit of  the  oscillator.  The  rate  at  which  these 
pulses  occur  is  known  as  the  pulse-repetition 
frequency,  or  p.r.f. 


10-5  Counting  Circuits 


A counting  circuit,  or  frequency  divider  is 
one  which  receives  uniform  pulses,  represent- 


Flgure  13 

STEP-BY-STEP  COUNTING  CIRCUIT 


ing  units  to  be  counted,  and  produces  a volt- 
age that  is  proportional  to  the  frequency  of 
the  pulses.  A counting  circuit  may  be  used  in 
conjunction  with  a blocking  oscillator  to  pro- 
duce a trigger  pulse  which  is  a submultiple  of 
of  the  frequency  of  the  applied  pulse.  Either 
positive  or  negative  pulses  may  be  counted. 
A positive  counting  circuit  is  shown  in  figure 
12A,  and  a negative  counting  circuit  is  shown 
in  figure  12B.  The  positive  counter  allows  a 
certain  amount  of  current  to  flow  through  R, 
each  time  a pulse  is  applied  to  C,. 

The  positive  pulse  charges  C„  and  makes 
the  plate  of  V2  positive  with  respect  to  its 
cathode.  V,  conducts  until  the  exciting  pulse 
passes.  Cl  is  then  discharged  by  V„  and  the 
circuit  is  ready  to  accept  another  pulse.  The 
average  current  flowing  through  Rj  increases 
as  the  pulse-repetition  frequency  increases, 
and  decreases  as  the  p.r.f.  decreases. 

By  reversing  the  diode  connections,  as 
shown  in  figure  12B,  the  circuit  is  made  to 
respond  to  negative  pulses.  In  this  circuit,  an 
increase  in  the  p.r.f.  causes  a decrease  in  the 
average  current  flowing  through  Rj,  which  is 
opposite  to  the  effect  in  the  positive  counter. 


Figure  14 

The  step-by-step  counter  used  to  trigger  a 
blocking  oscillator.  The  blocking  oscillator 
serves  as  a frequency  divider. 


HANDBOOK 


R-C  Oscillators  191 


Rl  X Cl  =R2  X C2 

Figure  15 

THE  WIEN-BRIDGE  AUDIO  OSCILLATOR 


Figure  16 

THE  PHASE-SHIFT  OSCILLATOR 


A step-counter  is  similar  to  the  circuits 
discussed,  except  that  a capacitor  which  is 
large  compared  to  C,  replaces  the  diode  load 
resistor.  The  charge  of  this  condenser  is  in- 
creased during  the  time  of  each  pulse,  pro- 
ducing a step  voltage  across  the  output  (figure 
13).  A blocking  oscillator  may  be  connected 
to  a step-counter,  as  shown  in  figure  14.  The 
oscillator  is  triggered  into  operation  when  the 
voltage  across  Cj  reaches  a point  sufficiently 
positive  to  raise  the  grid  of  V,  above  cutoff. 
Circuit  parameters  may  be  chosen  so  that  a 
count  division  up  to  1/20  may  be  obtained 
with  reliability. 


10-6  Resistance-Capacity 
Oscillators 

In  an  R-C  oscillator,  the  frequency  is  de- 
termined by  a resistance  capacity  network  that 
provides  regenerative  coupling  between  the 
output  and  input  of  a feedback  amplifier.  No 
use  is  made  of  a tank  circuit  consisting  of  in- 
ductance and  capacitance  to  control  the  fre- 
quency of  oscillation. 

The  Wien-Bridge  oscillator  employs  a V/ien 
network  in  the  R-C  feedback  circuit  and  is 
shown  in  figure  15.  Tube  Vi  is  the  oscillator 
tube,  and  tube  V,  is  an  amplifiet  and  phase- 
inverter  tube.  Since  the  feedback  voltage 
through  C4  produced  by  Vj  is  in  phase  with  the 
input  circuit  of  V,  at  all  frequencies,  oscilla- 
tion is  maintained  by  voltages  of  any  frequen- 
cy that  exist  in  the  circuit.  The  bridge  circuit 
is  used,  then,  to  eliminate  feedback  voltages 
of  all  frequencies  except  the  single  frequency 
desired  at  the  output  of  the  oscillator.  The 
bridge  allows  a voltage  of  only  one  frequency 
to  be  effective  in  the  circuit  because  of  the 
degeneration  and  phase  shift  provided  by  this 


circuit.  The  frequency  at  which  oscillation 
occurs  is: 

1 

f , when  R,  X Cl  = Rj  X C, 

2n  Rl  Cl 


A lamp  Lp  is  used  as  the  cathode  resistor 
of  Vi  as  a thermal  stabilizer  of  the  oscillator 
amplitude.  The  variation  of  the  resistance 
with  respect  to  current  of  the  lamp  bulb  holds 
the  oscillator  output  voltage  at  a nearly  con- 
stant amplitude. 

The  phase’Shift  oscillator  shown  in  figure 
16  is  a single  tube  oscillator  using  a three 
section  phase  shift  network.  Each  section  of 
the  network  produces  a phase  shift  in  propor- 
tion to  the  frequency  of  the  signal  that  passes 
through  it.  For  oscillations  to  be  produced, 
the  signal  from  the  plate  of  the  tube  must  be 
shifted  180®.  Three  successive  phase  shifts 
of  60°  accomplish  this,  and  the  frequency  of 
oscillation  is  determined  by  this  phase  shift. 
A high-mu  triode  or  a pentode  must  be  used  in 
this  circuit.  In  order  to  increase  the  frequency 
of  oscillation,  either  the  resistance  or  the 
capacitance  must  be  decreased. 


THE  BRIDGE-TYPE  PHASE-SHIFT 
OSCILLATOR 


192  Special  Vacuum  Tube  Circuits 


THE  RADIO 


IN  THE  HEATH  AG-9  AUDIO 
GENERATOR 


•NOTCH*  FREQUENCY 


ZITRC 

WHERE 

C=VCl  CE 


Figure  19 

BRIDGE-T  FEEDBACK 
LOOP  CIRCUITS 

Oscillation  will  occur  at  the  null 
frequency  of  the  bridge,  at  which 
frequency  the  bridge  allows 
minimum  degeneration  in  loop  2. 


A bridge-type  phase  shift  oscillator  is 
shown  in  figure  17.  The  bridge  is  so  propor- 
tioned that  at  only  one  frequency  is  the  phase 
shift  through  the  bridge  180°.  Voltages  of  other 
frequencies  are  fed  back  to  the  grid  of  the  tube 
out  of  phase  with  the  existing  grid  signal,  and 
are  cancelled  by  being  amplified  out  of  phase. 

The  NBS  Bridge-T  oscillator  developed  by 
the  National  Bureau  of  Standards  consists  of 
a two  stage  amplifier  having  two  feedback 
loops,  as  shown  in  figure  18.  Loop  1 consists 
of  a regenerative  cathode-to-cathode  loop,  con- 
sisting of  Lpi  and  C3,  The  bulb  regulates  the 
positive  feedback,  and  tends  to  stabilize  the 
output  of  the  oscillator,  much  as  in  the  man- 
ner of  the  Wien  circuit.  Loop  2 consists  of  a 
grid-cathode  degenerative  circuit,  containing 
the  bridge-T.  Oscillation  will  occur  at  the 
null  frequency  of  the  bridge,  at  which  frequen- 
cy the  bridge  allows  minimum  degeneration 
in  loop  2 (figure  19)- 


10-7  Feedback 

Feedback  amplifiers  have  been  d i s c u s s e d 
in  Chapter  6,  section  15  of  this  Handbook.  A 
more  general  use  of  feedback  is  in  automatic 
control  and  regulating  systems.  Mechanical 
feedback  has  been  used  for  many  years  in  such 
forms  as  engine  speed  governors  and  steering 
servo  engines  on  ships. 

A simple  feeTiback  system  for  temperature 
control  is  shown  in  figure  20.  This  is  a cause 


and  effect  system.  The  furnace  (F)  raises  the 
room  temperature  (T)  to  a predetermined  value 
at  which  point  the  sensing  thermostat  (TH) 
reduces  the  fuel  flow  to  the  furnace.  When  the 
room  temperature  drops  below  the  predeter- 
mined value  the  fuel  flow  is  increased  by  the 
thermostat  control.  An  interdependent  control 
system  is  created  by  this  arrangement:  the 
room  temperature  depends  upon  the  thermostat 
action,  and  the  thermostat  action  depends  upon 
the  room  temperature.  This  sequence  of  events 
may  be  termed  a closed  loop  feedback  system. 


Figure  20 

SIMPLE  CLOSED  LOOP 
FEEDBACK  SYSTEM 

Room  temperature  (T)  controls 
fuel  supply  to  furnace  (F)by  feed- 
back loop  through  Thermostat 
(TH)  control. 


HANDBOOK 


Feedback  193 


Figure  21 

PHASE  SHIFT  OF  ERROR 
SIGNAL  MAY  CAUSE  OSCILLA- 
TION INCLOSED  LOOP  SYSTEM 

To  prevent  oscillafiont  the  gain 
of  the  feedback  loop  must  be 
less  than  unity  when  the  phase 
shift  of  the  system  reaches  ISO 
degrees. 


Error  Cancellation  A feedback  control  system 
is  dependent  upon  a degree 
of  error  in  the  output  signal,  since  this  error 
component  is  used  to  bring  about  the  correc- 
tion. This  component  is  called  the  error  signal. 
The  error,  or  deviation  from  the  desired  signal 


is  passed  through  the  feedback  loop  to  cause 
an  adjustment  to  reduce  the  value  of  the  error 
signal.  Care  must  be  taken  in  the  design  of 
the  feedback  loop  to  reduce  over-control  ten- 
dencies wherein  the  correction  signal  would 
carry  the  sytem  past  the  point  of  correct  oper- 
ation. Under  certain  circumstances  the  new 
error  signal  would  cause  the  feedback  control 
to  overcorrect  in  the  opposite  direction,  result- 
ing in  hunting  or  oscillation  of  the  closed 
loop  system  about  the  correct  operating  point. 

Negative  feedback  control  would  tend  to 
dampout  spurious  system  oscillation  if  it  were 
not  for  the  time  lag  or  phase  shift  in  the  sys- 
tem. If  the  overall  phase  shift  is  equal  to  one- 
half  cycle  of  the  operating  frequency  of  the 
system  the  feedback  will  maintain  a steady 
state  of  oscillation  when  the  circuit  gain  is 
sufficiently  high,  as  shown  in  figure  21.  In 
order  to  prevent  oscillation,  the  gain  figure  of 
the  feedback  loop  must  be  less  than  unity  when 
the  phase  shift  of  the  system  reaches  180  de- 
grees. In  an  ideal  control  system  the  gain  of 
the  loop  would  be  constant  throughout  the 
operating  range  of  the  device,  and  would  drop 
rapidly  outside  the  range  to  reduce  the  band- 
width of  the  control  system  to  a minimum. 

The  time  lag  in  a closed  loop  system  may 
be  reduced  by  using  electronic  circuits  in  place 
of  mechanical  devices,  or  by  the  use  of  special 
circuit  elements  having  a phase-lead  charac- 
teristic. Such  devices  make  use  of  the  proper- 
ties of  a capacitor,  wherein  the  current  leads 
the  voltage  applied  to  it. 


CHAPTER  ELEVEN 


Electronic  Computers 


Mechanical  computing  machines  were  first 
produced  in  the  seventeenth  century  in  Eu- 
rope although  the  simple  Chinese  abacus  (a 
digital  computer)  had  been  in  use  for  cen- 
turies. Until  the  last  decade  only  simple  me- 
chanical computers  (such  as  adding  and  book- 
keeping machines)  were  in  general  use. 

The  transformation  and  transmission  of  the 
volume  of  information  required  by  modern 


technology  requires  that  machines  assume  many 
of  the  information  processing  systems  former- 
ly done  by  the  human  mind-  Computing  ma- 
chines can  perform  routine  operations  mote 
quickly  and  more  accurately  than  a human  be- 
ing, processing  mathematical  and  logistical 
data  on  a production  line  basis.  The  computer, 
however,  cannot  create,  but  can  only  follow 
instructions.  If  the  instructions  are  in  error. 


THE  IBM 
COMPUTER 
AND 

"MEMORY" 

The  "704"  Com- 
puter is  used  with 
32,0  00  "word" 
memory  storage 
unit  for  research 
programs.  Heart 
of  this  auxiliary 
unit  are  small, 
doughnut-shaped 
iron  ferrites  which 
store  information 
by  means  of  mag- 
netism. The  unit 
is  the  first  of  Hs 
kind  to  be  instal- 
led with  / B M's 
704  computer. 
Components  of  the 
system  seen  in  the 
foreground  are 
(left)  card  punch 
and  fright)  card 
reader.  In  the 
center  of  the  pic- 
ture is  the  704's 
processing  unit. 


Digital  Computers  195 


NORTH  SHOffE 


’•FARMER"  'CORN*  'HEN*  "FOX* 


SOUTH  SHORE 


NOTE:  ALL  BUTTONS  HAVE  ONE  NORMALLY  OPEN  CONTACT  AND 
ONE  NORMALLY  CLOSED  CONTACT. 

Figure  2 

A SEQUENCE  COMPUTER. 

Three  correct  buttons  will  sound  the  buzzer. 


Figure  1 

SIMPLE  PUZZLES  IN  LOGIC  MAY  BE 
SOLVED  BY  ELECTRIC  COMPUTER. 
THE  "FARMER  AND  RIVER" 
COMPUTER  IS  SHOWN  HERE. 


the  computer  will  produce  a wrong  answer. 

Computers  may  be  divided  into  two  classes: 
the  digital  and  the  analog.  The  digital  com- 
puter counts,  and  its  accuracy  is  limited  only  by 
the  number  of  significant  figures  provided  for 
in  the  instrument.  The  analog  c omputer 
measures,  and  its  accuracy  is  limited  by  the 
percentage  errors  of  the  devices  used,  multi- 
plied by  the  range  of  the  variables  they  repre- 
sent. 

11-1  Digital  Computers 

The  digital  computer  operates  in  discrete 
steps.  In  general,  the  mathematical  operations 
are  performed  by  combinations  of  additions. 
Thus  multiplication  is  performed  by  repeated 
additions,  and  integration  is  performed  by 
summation.  The  digital  computet  may  be 
thought  of  as  an  "on-off”  device  operating 
from  signals  that  either  exist  or  do  not  exist. 
The  common  adding  machine  is  a simple  com- 
puter of  this  type.  The  "on-off”  or  "yes-no” 
type  of  situation  is  well  suited  to  switches,  elec- 
trical relays,  or  to  electronic  tubes. 

A simple  electrical  digital  computer  may  be 
used  to  solve  the  old  “farmer  and  river”  prob- 
lem. The  farmer  must  transport  a hen,  a bushel 
of  corn,  and  a fox  across  a river  in  a small 
boat  capable  of  carrying  the  farmer  plus  one 
other  article.  If  the  farmer  takes  the  fox  in 
the  boat  with  him,  the  hen  will  eat  the  corn. 
On  the  other  hand,  if  he  takes  the  corn,  the 
fox  will  eat  the  hen.  The  circuit  for  a simple 
computer  to  solve  this  problem  is  shown  in 
figure  1.  When  the  switches  are  moved  from 
"south  shore”  to  "north  shore”  in  the  proper 
sequence  the  warning  buzzer  will  not  sound- 
An  error  of  choice  will  sound  the  buzzer. 

A second  simple  "digital  computer”  is  shown 
is  figure  2.  The  problem  is  to  find  the  three 


proper  push  buttons  that  will  sound  the  buzzer. 
The  nine  buttons  are  mounted  on  a board  so 
that  the  wiring  cannot  be  seen. 

Each  switch  of  these  simple  computers  ex- 
ecutes an  "on-off”  action.  When  applied  to  a 
logical  problem  "yes-no”  may  be  substituted 
for  this  term.  The  computer  thus  can  act  out 
a logical  concept  concerned  with  a simple 
choice.  An  electronic  switch  (tube)  may  be 
substituted  for  the  mechanical  switch  to  in- 
crease the  speed  of  the  computer.  The  early 
computers,  such  as  the  ENIAC  (Electronic  Nu- 
merical Integrator  and  Calculator)  employed 
over  18,000  mbes  for  memory  and  registering 
circuits  capable  of  "remembering”  a 10-digit 
number. 

11-2  Binary  Notation 

To  simply  and  reduce  the  cost  of  the  digital 
computer  it  was  necessary  to  modify  the  system 
of  operation  so  that  fewer  tubes  were  used  per 
bit  of  information.  The  ENlAC-type  computer 
requires  50  tubes  to  register  a 5-digit  number. 


© 

© 

© 

© 

© 

® 

® 

® 

@ 

® 

® 

® 

® 

® 

® 

® 

® 

® 

® 

® 

® 

® 

® 

® 

® 

® 

® 

© 

@ 

® 

® 

® 

® 

® 

® 

® 

® 

® 

® 

® 

® 

Figure  3 

BINARY  NOTATION  MAY  BE  USED 
FOR  DIGITAL  DISPLAY.  BINARY 
BOARD  ABOVE  INDICATES  "73092." 


196  Electronic  Computers 


THE  RADIO 


DIGIT 

TUBE(S) 

1 

1 

2 

2 

3 

2 + 1 

4 

4 

5 

4 + 1 

6 

4 + 2 

7 

4+2  + 1 

6 

a 

9 

a + 1 

10 

a + 2 

1 

a + 2+1 

12 

B + 4 

13 

a+4+ 1 

14 

a+4  + 2 

a + 4-h2  + 1 

Figure  4 

BINARY  DECIMAL  NOTATION.  ONLY 
FOUR  TUBES  ARE  REQUIRED  TO 
REPRESENT  DIGITS  FROM  1 TO  15. 
THE  DIGIT  "12"  IS  INDICATED 
ABOVE. 


The  tubes  (or  their  indicator  lamps)  can  be 
arranged  in  five  columns  of  10  tubes  each. 
From  right  to  left  the  columns  represent  units, 
tens,  hundreds,  thousands,  etc.  The  bottom  tube 
in  each  column  represents  "zero,”  the  second 
tube  represents  "one,”  the  third  tube  "two,” 
and  so  on.  Only  one  tube  in  each  column  is 
excited  at  any  given  instant.  If  the  number 
73092  is  to  be  displayed,  tube  number  seven 
in  the  fifth  column  is  excited,  tube  number 
three  in  the  fourth  column,  tube  number  zero 
in  the  third  column,  etc.  as  shown  in  figure  3. 

A simpler  system  employs  the  binary  deci- 
mal notation,  wherein  any  number  from  one 
to  fifteen  can  be  represented  by  four  tubes. 
Each  of  the  four  tubes  has  a numerical  value 
that  is  associated  with  its  position  in  the  tube 
group.  More  than  one  tube  of  the  group  may 
be  excited  at  once,  as  illustrated  in  figure  4. 
The  values  assigned  to  the  tubes  in  this  par- 
ticular group  are  1,  2,  4,  and  8.  Additional 
tubes  may  be  added  to  the  group,  doubling  the 
notation  of  the  mbe  thus:  1,  2,  4,  8,  I6,  32, 
64,  128,  356,  etc.  Any  numerical  value  lower 
than  the  highest  group  number  can  be  dis- 
played by  the  correct  tube  combination. 

A third  system  employs  the  binary  notation 
which  makes  use  of  a bit  (binary  digit)  repre- 
senting a single  morsel  of  information.  The 
binary  system  has  been  known  for  over  forty 
centuries,  and  was  considered  a mystical  revel- 
ation for  ages  since  it  employed  only  two  sym- 


DECIMAL  NOTATION 

BINARY  NOTATION 

0 

0 

1 

1 

2 

1,  0 

3 

1 , 1 

4 

1,  0,  0 

, 

1,  0, 1 

1. 1,  0 

1, 1, 1 

a 

1,  0,0.0 

ft 

1,  0,  0, 1 

10 

1,  0, 1.  0 

Figure  5 

BINARY  NOTATION  SYSTEM 
REQUIRES  ONLY  TWO  NUMBERS, 
"0"  AND  "1." 


bols  for  all  numbers.  Computer  service  usual- 
ly employs  "zero”  and  "one”  as  these  symbols. 
Decimal  notation  and  binary  notation  for  com- 
mon numbers  are  shown  in  figure  5.  The 
binary  notation  represents  4-digit  numbers 
(thousands)  with  ten  bits,  and  7-digit  num- 
bers (millions)  with  20  bits.  Only  one  elec- 
tron tube  is  required  to  display  an  information 
bit.  The  savings  in  components  and  primary 
power  drain  of  a binary-type  computer  over 
the  older  ENIAC-type  computer  is  obvious. 
Figure  6 illustrates  a computer  board  show- 
ing the  binary  indications  from  one  to  ten. 

Dig'i'ol  The  digital  computer  is  em- 

Computer  Uses  ployed  in  a "yes-no”  situa- 
tion. It  may  be  used  for 
routine  calculations  that  would  ordinarily  re- 
quire enormous  man-hours  of  time,  such  as 
checking  stress  estimates  in  aircraft  design,  or 
military  logistics,  and  problems  involving  the 
manipulation  of  large  masses  of  figures. 


DECIMAL  NOTATION 

COMPUTER  NOTATION 

0 

• 

• 

• 

• 

1 

• 

• 

• 

0 

2 

• 

• 

0 

• 

3 

• 

• 

0 

0 

4 

• 

0 

• 

• 

5 

• 

0 

• 

0 

e 

• 

0 

0 

• 

• 

0 

0 

0 

a 

0 

• 

• 

• 

9 

0 

• 

• 

0 

10 

0 

• 

0 

• 

• = 

OFF 

0 = 

ON 

Figure  6 

BINARY  NOTATION  AS  REPRESENTED 
ON  COMPUTER  BOARD  FOR  NUMBERS 
FROM  1 TO  10. 


HANDBOOK 


Analog  Computers  197 


eour=ei+e2 


Figure  7 

SUMMATION  OF  TWO  VOLTAGES 
BY  ELECTRICAL  MEANS. 


11-3  Analog  Computers 

The  analog  computer  represents  the  use  of 
one  physical  system  as  a model  for  a second 
system  that  is  usually  more  difficult  to  con- 
struct or  to  measure,  and  that  obeys  the  equa- 
tions of  the  same  form.  The  term  analog  im- 
plies similarity  of  relations  or  properties  be- 
tween the  two  systems.  The  common  slide-rule 
is  a mechanical  analog  computer.  The  speedo- 
meter in  an  automobile  is  a differential  analog 
computer,  displaying  information  proportional 
to  the  rate  of  change  of  speed  of  the  vehicle. 
The  electronic  analog  computer  employs  cir- 
cuits containing  resistance,  capacitance,  and  in- 
ductance arranged  to  behave  in  accordance  with 
analogous  equations.  Variables  are  represented 
by  d-c  voltages  which  may  vary  with  time. 


+ 150V. 


Figure  8 

SUMMATION  OF  TWO  VOLTAGES 
BY  ELECTRONIC  MEANS. 


Thus  complicated  problems  can  be  solved  by 
d-c  amplifiers  and  potentiometer  controls  in 
electronic  circuits  performing  mathematical 
functions. 

AddiKon  and  If  a linear  network  is  ener- 
Subtroction  gized  by  two  voltage  sources 
the  voltages  may  be  summed 
as  shown  in  figure  7.  Subtraction  of  quantities 
may  be  accomplished  by  using  negative  and 
positive  voltages.  A-c  voltages  may  be  em- 
ployed for  certain  additive  circuits,  and  more 


THE 

HEATHKIT 

ELECTRONIC 

ANALOG 

DIGITAL 

COMPUTER 

This  **  electronic 

slide  rule*'  s/mu- 
lates  equations  or 
physical  problems 
electronically,  sub- 
stituting one  phys- 
ical system  as  a 
model  for  a sec- 
ond system  that 
is  usually  more 
difficult  or  costly 
fo  construct  or 
measure,  and  that 
obeys  equations  of 
the  same  form. 


198  Electronic  Computers 


THE  RADIO 


Figure  9 

ELECTRONIC  MULTIPLICATION 
MAY  BE  ACCOMPLISHED  BY 
CALIBRATED  POTENTIOMETERS, 
WHEN  OUTPUT  VOLTAGE  IS 
PROPORTIONAL  TO  THE  INPUT 
VOLTAGE  MULTIPLIED  BY  A 
CONSTANT  (R2/R1). 

complex  circuits  employ  vacuum  tubes,  as  in 
figure  8.  Synchronous  transformers  may  be 
used  to  add  expressions  of  angular  rotation, 
and  circuits  have  been  developed  for  adding 
time  delays,  or  pulse  counting. 

MultiplicaHon  Electronic  multiplication  and 

and  Division  division  may  be  accomplished 

with  the  use  of  potentiome- 
ters where  the  output  voltage  is  proportional  to 
the  input  voltage  multiplied  by  a constant 
which  may  be  altered  by  changing  the  physical 
arrangement  of  the  potentiometer  (figure  9). 
Variable  autotransformers  may  also  be  used  to 
perform  multiplication. 

A simple  bridge  may  be  used  to  obtain  an 
output  that  is  the  product  of  two  inputs  divided 
by  a third  input,  as  shown  in  figure  10. 

Differentiation  The  time  derivative  of  a 
voltage  can  be  expressed  as 
a charge  on  a capacitor  by; 

. _ P de 

‘ ^dt  (1) 

and  is  shown  in  figure  11  A.  The  charging  cur- 
rent is  converted  into  a voltage  by  the  use 
of  a resistor,  R.  If  the  input  to  the  RC  circuit 
is  charging  at  a uniform  rate  so  that  the  cur- 
rent through  C and  R is  constant,  the  output 
voltage  Co  is; 


Figure  1 1 

ELECTRONIC  DIFFERENTIATION 

The  time  derivatiYe  of  a voltage  can  be 
expressed  as  a charge  on  a capacitor  (A). 
Operational  amplifier  (B)  employs  feed 
back  principle  for  short  differentiation 
time. 


dt  (2) 

For  highest  accuracy,  a small  RC  product 
should  be  used,  permitting  the  maximum  pos- 
sible differentiation  time.  The  output  of  the 
differentiator  may  be  amplified  to  any  suitable 
level. 

A more  accurate  differentiating  device 
makes  use  of  an  operational  amplifier.  This 
unit  is  a high  gain,  negative  feedback  d-c 
amplifier  (discussed  in  section  11-4)  with  the 
resistance  portion  of  the  RC  product  appearing 
in  the  feedback  loop  of  the  amplifier  (figure 
IIB).  A shorter  differentiation  time  may  be 
employed  if  the  junction  point  between  R and 
C could  be  held  at  a constant  potential.  The 
feedback  amplifier  shown  inverts  the  output 
signal  and  applies  it  to  the  RC  network,  hold- 
ing the  junction  potential  constant. 

Integration  Integration  is  a process  of  ac- 
cumulation, or  summation,  and 
requires  a device  capable  of  storing  physical 
quantities.  A capacitor  will  store  an  electrical 
charge  and  will  give  the  time  integral  of  a 
current  in  respect  to  a voltage; 

Co  = — i dt  ( 3 ) 

c 

In  most  computers,  the  input  signal  is  in 


Figure  1 0 

ELECTRONIC  MULTIPLICATION  BY 
BRIDGE  CIRCUIT  PROVIDES 
OUTPUT  THAT  IS  PRODUCT  OF 
TWO  INPUTS  DIVIDED  BY  A 
THIRD  INPUT. 


the  form  of  a voltage,  and  the  input  charging 
current  of  the  capacitor  must  be  taken  through 
a series  resistance  as  in  figure  12.  If  the  inte- 
grating time  is  short  the  charging  current  is 
approximately  proportional  to  the  input  vol- 
tage. The  charging  current  may  be  made  a 
true  measure  of  the  input  voltage  by  the  use 

Figure  12 
SIMPLE 
INTEGRATION 
CIRCUIT 

Making  use  of 
charging  current 
of  capacitor. 


HANDBOOK 


Operational  Amplifier  199 


eouT 


Figure  13 

"MILLER  FEEDBACK"  INTEGRATOR 
SUITABLE  FOR  COMPUTER  USE. 


eouT 


Figure  14 

R-L  NETWORK  USED  FOR 
INTEGRATION  PURPOSES. 


Rf 


eo 


Figure  15 

OPERATIONAL  AMPLIFIER  (-A) 

Mathematical  operations  may  be  performed 
by  any  operational  amplifier,  usually  a 
stable,  high^gain  d-e  amplifier,  such  as 
shown  in  Figure  16. 

of  an  operational  amplifier  wherein  the  ca- 
pacitance portion  of  the  RC  product  appears 
in  the  feedback  loop  of  the  amplifier,  holding 
the  junction  point  between  R and  C at  a con- 
stant potential.  A simple  integrator  is  shown 
in  figure  13  employing  the  Miller  feedback 
principle.  Integration  is  also  possible  with  an 
RL  network  (figure  14). 


11-4  The  Operational 
Amplifier 

Mathematical  operations  are  performed  by 
using  a high  gain  d-c  amplifier,  termed  an 
operational  amplifier-  The  symbol  of  this  unit 
is  a triangle,  with  the  apex  pointing  in  the  di- 
rection of  operation  (figure  15).  The  gain  of 
such  an  amplifier  is  A,  so: 


eo  — — Aeg,  or  eg  — — -7-  , , , 

A (4) 

If  A approaches  infinity,  eg  will  be  ap- 
proximately zero.  In  practice  this  condition  is 
realized  by  using  amplifiers  having  open  loop 
gains  of  30,000  to  60,000.  If  eo  is  set  at  100 
volts,  eg  will  be  of  the  order  of  a few  milli- 
volts. Thus,  considering  eg  equal  to  zero: 


which  may  be  written: 


eo  = ~ Kei,  where  K = ,,, 

Ri  (6) 


This  amounts  to  multiplication  by  a constant 
coefficient,  since  Ri  and  Rr  may  be  fixed  in 
value.  The  circuit  of  a typical  operational  am- 
plifier is  shown  in  figure  16. 

Amplifier  Two  voltages  may  be  added  by 
Operation  the  amplifier,  as  shown  in  figure 
17.  Keeping  in  mind  that  eg  is 


200  Electronic  Computers 


THE  RADIO 


Rf 

1— iW 1 

Ri 

’ ^ ^ 

1 ‘ 
ec  9 

Figure  17 

TWO  VOLTAGES  MAY  BE 
ADDED  BY  SUMMATION 
AMPLIFIER. 


essentially  at  zero  (ground)  potential; 


Co  — ~ 


Ri 


e2  + 


(7) 


Figure  19 
"MASS-SPRING- 
DAMPER"  PROBLEM 
MAY  BE  SOLVED  BY 
ELECTRICAL  ANALOGY 
WITH  SIMPLE 
COMPUTER. 


or,  eo  — ■ 


K.  ft  -f  K2  e2 


where  Ki  — — — and  K3  — ^ 
Ri  R2 


(8) 

(9) 


As  long  as  eo  does  nor  exceed  the  input 
range  of  the  amplifier,  any  number  of  inputs 
may  be  used; 


Rf  . Rf  I I 

eo  = - — ft  + -r-ft+ - - - en 

ixi  IV2  i\n 


(10) 


or 


Co  — Rt  + 


R„ 


(11) 


Integration  is  performed  by  replacing  the 
feedback  resistor  Rt  with  a capacitor  Ct,  as 
shown  in  figure  18.  For  rhis  circuit  (with  eg 
approximately  zero ) ; 

. dg  ei 

^ (12) 


L r'  dg  dec  , ei  _ dec 

but  g— Cf  Co,  so— I Cf^ — and  — Cf  -] — 

dt  dt  Ri  at 


(13) 

Thus;  deo  — ^ ei  dt 

Ri  Cf 

(14) 

and  eo  = ^ ei  dt 

Ri  Cf 

(15) 

Figure  18 
INTEGRATION 

Performed  by 
Summation  Amplifier 
by  replacing  feed 
back  resistor  with  a 
capacitor. 


11-5  Solving  Analog 
Problems 


By  combining  the  above  operations  in  var- 
ious ways,  problems  of  many  kinds  may  be 
solved  For  example,  consider  the  mass-spring- 
damper  assembly  shown  in  figure  19.  The  mass 
Af  is  connected  to  the  spring  which  has  an 
elastic  constant  K.  The  viscous  damping  con- 
stant is  C.  The  vertical  displacement  is  y.  The 
sum  of  the  forces  acting  on  mass  M is; 

f (t)  =M^  + C^  + Ky  (16) 

where  f (t)  is  the  applied  force,  or  forcing 
function. 

The  first  step  is  to  set  up  the  analog  com- 
puter circuit  so  as  to  obtain  an  output  voltage 
proportional  to  y for  a given  input  voltage 
proportional  to  f ( t ) . Equation  (16)  may  be 
rewritten  in  the  form; 

(17) 


If,  in  the  analog  circuit,  there  is  a voltage 

equal  to  M -7^  it  can  be  converted  to  — 

dt"  dt 

by  passing  it  through  an  integrator  circuit  hav- 
ing an  RC  time  constant  equal  to  M.  This  re- 
sulting voltage  can  be  passed  through  a second 
integrator  stage  with  unit  time  constant  which 
will  have  an  output  volage  equal  to  y.  The 
dy 

voltages  representing  y,  — > and  f (t)  can 
dt 


dy 

then  be  summed  to  give  C Ky  + f (t) 

dt 


which  is  the  right  hand  side  of  equation  (17), 

d"y 

and  therefore  equal  to  M ^-5- . Connecting  the 
dt 


HANDBOOK 


Analog  Problems  201 


Figure  20 

ANALOG  SOLUTION  FOR  "MASS-SPRING- 
DAMPER"  PROBLEM  OF  FIGURE  19. 


output  of  the  summing  amplifier  (A3)  to  the 
input  of  the  first  as  shown  in  figure  20  satis- 
fies the  equation. 

To  obtain  a solution  to  the  problem,  the 
initial  displacement  and  velocity  must  be  speci- 
fied. This  is  done  by  charging  the  integrating 
capacitors  to  the  proper  voltages.  Three  oper- 
ational amplifiers  and  a summing  amplifier 
are  required. 

A second  problem  that  may  be  solved  by 
the  analog  computet  is  the  example  of  a freely 
falling  body.  Disregarding  air  resistance,  the 
body  will  fall  (due  to  the  action  of  gravity) 
with  a constant  acceleration.  The  equation 
describing  this  action  is: 


p - - 'i’v 

F — mg  — m 


(18) 


Integration  of  equation  (18)  will  give  the 

velocity,  or  , and  integration  a second  time 
dt 

will  give  displacement,  or  y.  The  block  dia- 
gram of  a suitable  computer  for  this  problem 
is  shown  in  figure  21. 

If  a voltage  proportional  to  g and  hence  to 

dV  , 


dt“ 


is  introduced  into  the  first  amplifier,  the 

dy 

output  of  that  unit  will  be  , or  the 


Figure  21 

ANALOG  COMPUTER  FOR 
"FREELY  FALLING  BODY" 
PROBLEM. 


velocity.  That,  in  turn,  will  become  y,  or  dis- 
tance, at  the  output  of  the  second  amplifier. 

Before  the  problem  can  be  solved  on  the 
computer  it  is  necessary  to  determine  the  time 
of  solution  desirable  and  the  output  amplitude 
of  the  solution.  The  time  of  solution  is  de- 
termined by  the  RC  constant  of  the  integrating 
amplifiers.  If  RC  is  set  at  unity,  computet 
time  is  equal  to  real  time.  The  computer  time 
desirable  is  determined  by  the  method  of  read- 
out. When  using  an  oscilloscope  for  read-out, 
a short  solution  time  is  desirable.  For  a re- 
corder, longer  solution  time  is  better. 

Suppose,  for  example,  in  the  problem  of  the 
falling  body,  the  distance  of  fall  in  2.5  seconds 
is  desired.  Using  an  RC  constant  of  1 would 
give  a solution  time  of  2.5  seconds.  This  would 
be  acceptable  for  a recorder  but  is  slow  for  an 
oscilloscope.  A convenient  time  of  solution  for 
the  ’scope  would  be  25  milliseconds.  This  is 
1/100  of  the  real  time,  so  an  RC  constant  of 
.01  is  needed.  This  can  be  obtained  with  C 
equal  to  0.1  gfd,  and  R equal  to  100,000 
ohms. 

It  is  now  necessary  to  choose  an  input  volt- 
age which  will  not  overdrive  the  amplifiers. 
The  value  of  g is  known  to  be  approximately 
32  ft/ sec. /sec.  A check  indicates  that  if  we 
set  g equal  to  32  volts,  the  voltage  represent- 
ing the  answer  will  exceed  100  volts.  Since 
the  linear  response  of  the  amplifier  is  only 
100  volts,  this  is  undesirable.  An  input  of  16 
volts,  however,  should  permit  satisfactory  op- 
eration of  the  amplifiers.  Output  voltages  near 
zero  should  also  be  avoided.  In  general,  output 
voltage  should  be  about  50  volts  or  so,  with 
amplifier  gains  of  20  to  60  being  preferable- 
Thus,  for  this  particular  problem  the  time- 
scale  factor  and  amplitude-scale  factor  have 


202  Electronic  Computers 


THE  RADIO 


Figure  22 

ANALOG  SOLUTION  FOR 
"FALLING  BODY  PROBLEM" 


Figure  23 

READ-OUT  SOLUTION  OF  "FREELY 
FALLING  BODY"  PROBLEM. 


been  chosen.  The  problem  now  looks  like 
figure  22. 

To  solve  the  problem,  relays  A and  A'  are 
opened.  The  solution  should  now  appear  on 
the  oscilloscope  as  shown  in  figure  23-  The 
solution  of  the  problem  leaves  the  integrating 
capacitors  charged.  It  is  necessary  to  remove 
this  charge  before  the  problem  can  be  rerun. 
This  is  done  by  closing  relays  A and  A'. 


LIMITING  CIRCUIT  TO  SIMULATE 
NON-LINEAR  FUNCTIONS  SUCH  AS 
ENCOUNTERED  IN  HYSTERESIS, 
BACKLASH,  AND  FRICTION 
PROBLEMS. 


11-6  Non-linear  Functions 

Problems  are  frequently  encountered  in 
which  non-linear  functions  must  be  simulated. 
Non-linear  potentiometers  may  be  used  to  sup- 
ply an  unusual  voltage  source,  or  diodes  may 
be  used  as  limiters  in  those  problems  in  which 
a function  is  defined  differently  for  different 
regions  of  the  independent  variable.  Such  a 
function  might  be  defined  as  follows; 

Go  — , ei  Ki  (19) 

eo~ei,  K ei  K2  (20) 

e©  — K2 , ei  K2  (21) 

where  Ki  and  K2  are  constants. 

Various  limiting  circuits  can  be  used,  one  of 
which  is  shown  in  figure  24.  This  is  a series 
limiter  circuit  which  is  simple  and  does  not 
require  special  components.  Commonly  en- 
countered problems  requiring  these  or  similar 
limiting  techniques  include  hysteresis,  back- 
lash, and  certain  types  of  friction. 


SIMPLIFIED  DIAGRAM  OF  FUNCTIONAL  GENERATOR  TO  APPROXIMATE  NON-LINEAR 

FUNCTIONS. 


HANDBOOK 


Non-Linear  Functions  203 


Xo  Xl  Xz  X3  X4  X5  X6+X-» 

Figure  26 

TYPICAL  NON-LINEAR  FUNCTION 
WHICH  MAY  BE  SET  UP  WITH 
FUNCTION  GENERATOR. 


The  Function  A function  generator  may  be 
Generator  used  to  approximate  almost 

any  non-linear  function.  This 
is  done  by  use  of  straight  line  segments  which 
are  combined  to  approximate  curves  such  as 
are  found  in  trigonometric  functions  as  well 
as  in  stepped  functions.  In  a typical  generator 
ten  line  segments  are  used,  five  in  the  plus-x 
direction,  and  five  in  the  minus-x  direction. 
Five  6AL5  double  diodes  are  used.  Each  line 
segment  is  generated  by  a modified  bridge 
circuit  (figure  25).  A ramp  function  or  volt- 
age is  fed  into  one  arm  of  the  bridge  while 
the  opposite  atm  is  connected  to  a biased  diode. 
The  other  two  arms  of  the  bridge  combine  to 
form  the  output.  The  voltage  appearing  across 
one  of  these  arms  is  fed  through  a sign-chang- 
ing amplifier  and  then  summed  with  the  volt- 
age appearing  at  the  opposite  arm.  If  the  arm 
of  potentiometer  P (the  slope  control)  is  set 
in  the  center,  the  bridge  will  be  balanced  and 


Figure  27 

ELECTRONIC  PACKAGE  IN 
DIGITAL  COMPUTER. 


Stylized  diagram  of  tube  package.  Lines  car- 
rying negative  puises  are  marked  by  a smail 
circle  at  each  end.  Gates  are  indicated  by 
a semi-circle  with  "pins"  for  each  input. 


the  output  of  the  summing  amplifier  will  be 
zero.  If,  on  the  other  hand,  potentiometer  P 
is  adjusted  one  way  or  the  other  from  center, 
the  bridge  will  be  unbalanced  and  the  sum- 
ming amplifier  output  will  vary  linearly  with 
respect  to  the  input  in  either  a positive  or  neg- 
ative y direction,  depending  upon  which  side 
of  center  potentiometer  P is  set. 

The  breai  voltage,  or  value  of  x at  which  a 
straight  line  segment  will  begin  is  set  by 
biasing  the  diode  to  the  particular  voltage  level 
or  value  of  x desired.  The  ramp  function  gen- 
erator has  either  a positive  or  negative  input 
which  because  of  the  180  degree  phase  shift 
in  the  amplifier,  gives  a minus-  or  plus-x  out- 
put respectively.  A typical  function  such  as 
shown  in  figure  26  may  be  set  up  with  the 
function  generator.  The  initial  condition  volt- 


IBM's  new  "60S,"  the 
first  completely  tran- 
sistorized calculator 
for  commercial  appli- 
cations, operates 
without  the  use  of  a 
single  vacuum  tube. 
Transistors — tiny  ger- 
manium devices  that 
perform  many  of  the 
functions  of  conven- 
tional vacuum  tubes 
— moke  possible  50% 
reduction  in  compu- 
ter-unit size  and  a 
90%  reduction  in 
power  requirements 
over  a comparable 
IBM  tube-model  ma- 
chine. They  are 
mounted,  along  with 
related  circuitry,  on 
banks  of  printed  wir- 
ing panels  in  the 
608. 

The  machine's  inter- 
nal storage,  or  "mem- 
ory," is  made  up  of 
magnetic  cores — mi- 
nute. doughnut-shap- 
ed objects  that  can 
"remember"  informa- 
tion indefinitely,  and 
recall  it  for  use  in 
calculations  in  a few 
millionths  of  a 


204  Electronic  Computers 


THE  RADIO 


age  is  set  to  the  value  of  Xi.  The  break-voltage 
control  is  increased  until  the  output  of  the 
summing  amplifier  increases  abruptly,  indi- 
cating the  diode  is  conducting.  The  input 
voltage  from  the  initial  condition  power  supply 
is  set  to  the  value  X2  The  slope  control  (P) 
is  now  set  to  value  Y2.  A second  function  gen- 
erator may  be  used  to  set  points  Xa  and  X*, 
using  the  break-voltage  control  and  the  initial 
condition  voltage  adjustments.  Points  Xs  and 
Xo  are  finally  set  with  a third  generator.  The 
x-output  of  the  function  generator  system  may 
be  read  on  an  oscilloscope,  using  the  x-output 
of  the  ramp-function  generator  amplifier  as 
the  horizontal  sweep  for  the  oscilloscope. 

11-7  Digital  Circuitry 

Digital  circuits  dealing  with  "and,”  "or,” 
and  "not”  situations  may  be  excited  by  electri- 
cal pulses  representing  these  logical  operations. 
Sorting  and  amplifying  the  pulses  can  be  ac- 
complished by  the  use  of  electronic  packages, 
such  as  shown  in  figure  27.  Logical  operations 
may  be  accomplished  by  diode-resistor  gates 
operating  into  an  amplifier  stage.  Negative  and 
positive  output  pulses  from  the  amplifier  are 
obtained  through  diode  output  gates.  The  driv- 
ing pulses  may  be  obtained  from  a standard 
oscillator,  operating  at  or  near  1 me. 

A circuit  of  a single  digital  package  is  shown 
in  figure  28.  Other  configurations,  such  as  a 
"flip-flop"  may  be  used.  Many  such  packages 


can  be  connected  in  series  to  form  operational 
circuits.  The  input  "and”  and  "or”  gates  are 
biased  to  conduction  by  external  voltages.  The 
"and"  diode  gate  transmits  a pulse  only  when 
all  the  input  terminals  are  pulsed  positively, 
and  the  "or”  diode  gate  transmits  a positive 
pulse  applied  to  any  one  of  its  input  terminals. 
The  input  pulses  pass  through  the  gates  and 
drive  the  amplifier  stage,  which  delivers  an 
amplified  pulse  to  the  positive  and  negative 
output  gates,  and  to  accompanying  memory 
circuits. 

Memory  Circuits  A memory  circuit  consists 
of  some  sort  of  delay  line 
which  is  capable  of  holding  an  information 
pattern  for  a period  of  time.  A simple  electri- 
cal delay  line  capable  of  holding  a pulse  for 
a fraction  of  a microsecond  is  shown  in  figure 
29.  The  amount  of  delay  is  proportional  to 
the  frequency  of  the  input  signal  as  shown  in 
the  graph.  A "long”  transmission  line  may 
also  be  used  as  a delay  line,  with  the  signal 
being  removed  from  the  "far”  end  of  the  line 
after  being  delayed  an  interval  equal  to  the 
time  of  transmission  along  the  line.  Lines  of 
this  type  are  constructed  much  in  the  manner 
of  a standard  coaxial  cable,  except  that  the 
inner  conductor  is  a long,  thin  coil  as  illustrat- 
ed in  figure  30.  Other  memory  circuits  make 
use  of  magnetostrictive  or  piezoelectric  effects 
to  retard  the  pulse.  Information  may  also  be 
stored  in  electrostatic  storage  tubes,  upon  mag- 


HANDBOOK 


Digital  Circuitry  205 


I I I ' 


Figure  29 

SIMPLE  DELAY  LINE  COMPOSED  OF 
LUMPED  INDUCTANCE  AND 
CAPACITANCE.  GRAPH  SHOWS  HOW 
DELAY  TIME  VARIES  WITH 
FREQUENCY  OF  INPUT  PULSE. 


netic  recording  tape,  and  in  ferromagnetic 
cores.  Experimental  magnetic  cores  have  been 
developed  that  will  permit  construction  of 


Figure  30 

UNIFORM  DELAY  LINE  BEARS 
RESEMBLANCE  TO  COAXIAL  CABLE, 
EXCEPT  FOR  COILED  INNER 
CONDUCTOR. 


memories  capable  of  retaining  over  10,000 
separate  bits  of  information.  Tbe  memory  may 
be  "read"  or  ''interrogated”  by  sending  pulses 
through  the  matrix  and  observing  tbe  resulting 
induced  voltage  in  a secondary  circuit.  The 
output  from  the  memory  circuit  may  be  used 
to  excite  a program  control  system,  permitting 
the  computer  to  drive  another  mechanism  in 
the  proper  sequence. 


E 8.  E TECHNI-SHEET 

TABLE  OF  ALUMINUM  TUBING,  ROUND  DRAWN,  12  FOOT  LENGTHS 


3S-H14  TUBING 
DIAM.  WALL  |WT«/F 

3/16"  .022"  .013 

1/4"  .022"  .016 

1/4"  .035"  .028 

3/8"  .022"  .0  28 

3/8"  .035"  .060 

1/2"  .035"  .060 

1/2"  .049"  .083 

1/2"  .065"  .105 

5/8"  .035"  .076 

5/8"  .049"  .104 

5/8"  .065"  .133 

_3/^  J^^O 35  " 

3/4"  ,049"  .128 


24S-T3  TUBING  61S-T6  TUBING 


WALL 

WT«/FT. 

DIAM. 

WALL 

WT.«/FT 

DIAM. 

■022" 

.013 

1/4" 

.035" 

.028 

3/16" 

■022" 

.018 

3/8" 

■049" 

.060 

1/4" 

■035" 

■026 

1/2" 

■035" 

.061 

1/4" 

.022" 

■0  28 

5/8" 

.035" 

■078 

5/16" 

.035" 

■060 

3/4" 

.035" 

■094 

5/16" 

■035" 

■060 

3/4" 

■049" 

■ 130 

3/6" 

■049" 

■083 

7/8" 

.049" 

■151 

3/8" 

■06  5" 

■105 

1 " 

.035" 

.128 

3/8" 

.03  5" 

.076 

1 " 

.049" 

.175 

7/16" 

.049" 

■1  04 

1" 

■065" 

■228 

7/16" 

.065" 

.133 

1 1/4" 

.049" 

■221 

1/2" 

.035" 

.09  2 

1 1/2" 

.035" 

.193 

1/2" 

.049" 

.128 

1 1/2" 

.049" 

■268 

1/2" 

WALL 

.035" 
■035" 
■056" 
■035" 
■056" 
■0  35" 
■056" 
■06  5" 
.035" 
■065 
■028" 
^03^" 
.058 


WT.#/FT. 
.01  6 
■028 
■041 
■036 
■055 
■ 044 
■066 
■074 
.0  5 2 
■069 
.049 
^060 
.095 


■056"  .121 
■035'  .092 


1 " 

3S-H14  ALUMINUM  TUB  I N G - 05  A/ 1 1/4" 
B£A/r  OR  FLATTENED  READILY. 

24S-T3  ALUMINUM  TUBING-A'45  zV/G//  . 
TENSILE  STRENGTH  AND  CAN  BE  BENT  OR  ^ 
FLATTENED  ONLY  SLIGHTLY  WITHOUT  M/2" 

CRACKING.  GOOD  FOR  ANTENNA  MASTS,  1 1/2" 
BOOMS,  ELEMENTS  AND  OTHER  SECTIONS  2." 
REQUIRING  STRENGTH  AND  RIGIDITY. 

61S-T6  ALUMINUM  TUBING-//45  GREAT  — 

TENSILE  STRENGTH  BUT  CANNOT  BE  FLATTENED 


.035" 


.056" 


■083"  .2 

■058"  Ta 
■063"  .3 

■058"  .2 

■058"  .3 
■063"  .5 

■0  58"  .4 

■083^nT5 

OR  BENT. 


CHAPTER  TWELVE 


Radio  Receiver  Fundamentals 


A conventional  reproducing  device  such  as 
a loudspeaker  or  a pair  of  earphones  is  in- 
capable of  receiving  directly  the  intelligence 
carried  by  the  carrier  wave  of  a radio  trans- 
mitting station.  It  is  necessary  that  an  addi- 
tional device,  called  a radio  receiver,  be 
placed  between  the  receiving  antenna  and  the 
loudspeaker  or  headphones. 

Radio  receivers  vary  widely  in  their  com- 
plexity and  basic  design,  depending  upon  the 
intended  application  and  upon  economic  fac- 
tors. A simple  radio  receiver  for  reception  of 
radiotelephone  signals  can  consist  of  an  ear- 
phone, a silicon  or  germanium  crystal  as  a 
carrier  rectifier  or  demodulator,  and  a length 
of  wire  as  an  antenna.  However,  such  a re- 
ceiver is  highly  insensitive,  and  offers  no 
significant  discrimination  between  two  sig- 
nals in  the  same  portion  of  the  spectrum. 

On  the  other  hand,  a dual-diversity  receiver 
designed  for  single-sideband  reception  and 
employing  double  or  triple  detection  might 
occupy  several  relay  racks  and  would  cost 
many  thousands  of  dollars.  However,  conven- 
tional communications  receivers  are  interme- 
diate in  complexity  and  performance  between 
the  two  extremes.  This  chapter  is  devoted  to 
the  principles  underlying  the  operation  of 
such  conventional  communications  receivers. 


12-1  Detection  or 

DeiTiodulation 

A detector  or  demodulator  is  a device  for 
removing  the  modulation  (demodulating)  or 
detecting  the  intelligence  carried  by  an  in- 
coming radio  wave. 

Radiotelephony  Figure  1 illustrates  an  ele- 

Demodulation  mentary  form  of  radiotele- 

phony receiver  employing  a 
diode  detector.  Energy  from  a passing  radio 

wave  will  induce  a voltage  in  the  antenna  and 
cause  a radio-frequency  current  to  flow  from 
antenna  to  ground  through  coil  Lj.  The  alter- 
nating magnetic  field  set  up  around  Li  links 
with  the  turns  of  Lj  and  causes  an  r-f  current 
to  flow  through  the  parallel-tuned  circuit, 
Lj-Ci.  When  variable  capacitor  is  adjusted 
so  that  the  tuned  circuit  is  resonant  at  the 
frequency  of  the  applied  signal,  the  r-f  voltage 
is  maximum.  This  r-f  voltage  is  applied  to  the 
diode  detector  where  it  is  rectified  into  a vary- 
ing direct  current  and  passed  through  the  ear- 
phones. The  variations  in  this  current  corres- 
pond to  the  voice  modulation  placed  on  the 
signal  at  the  transmitter.  As  the  earphone 
diaphragms  vibrate  back  and  forth  in  accord- 


207 


208  Radio  Receiver  Fundamentals 


THE  RADIO 


Figure  1 

ELEMENTARY  FORM  OF  RECEIVER 

This  is  the  basis  of  the  **crystal  set**  type  of  re- 
ceiver, although  a vacuum  diode  may  be  used  in 
place  of  the  crystal  diode.  The  tank  circuit  L2~^i 
is  tuned  to  the  frequency  it  is  desired  to  receive. 
The  by-pass  capacitor  across  the  phones  should 
have  a low  reactance  to  the  carrier  frequency  be- 
ing received,  but  a high  reactance  to  the  modula- 
tion on  the  received  signal. 


PLATE-TICKLER  REGENERATION  WITH  "THROTTLE* 
CONDENSER  REGENERATION  CONTROL. 


QPklTr^f^c  Aiinin  oiixoiiT 


CATHODE-TAP  REGENERATION  WITH  SCREEN  VOLTAGE 
REGENERATION  CONTROL. 


ance  with  the  pulsating  current  they  audibly 
reproduce  the  modulation  which  was  placed 
upon  the  carrier  wave. 

The  operation  of  the  detector  circuit  is 
shown  graphically  above  the  detector  circuit 
in  figure  I.  The  modulated  carrier  is  shown 
at  A,  as  it  is  applied  to  the  antenna.  B repre- 
sents the  same  carrier,  increased  in  amplitude, 
as  it  appears  across  the  tuned  circuit.  In  C 
the  varying  d-c  output  from  the  detector  is 
seen. 


Figure  2 

REGENERATIVE  DETECTOR  CIRCUITS 
Regenerative  detectors  are  seldom  used  at  the  pre- 
sent  time  due  to  their  poor  selectivity.  However, 
they  do  Illustrate  the  simplest  type  of  receiver 
which  may  be  used  either  for  radiophone  or  radio- 
telegraph reception. 


Radiotelegraphy  Since  a c-w  telegraphy  sig- 
Reception  nal  consists  of  an  unmodu- 

lated carrier  which  is  inrer- 
rupted  to  form  dots  and  dashes,  it  is  apparent 
that  such  a signal  would  not  be  made  audible 
by  detection  alone.  While  the  keying  is  a form 
of  modulation,  it  is  composed  of  such  low  fre- 
quency components  that  the  keying  envelope 
itself  is  below  the  audible  range  for  hand  key- 
ing speeds.  Some  means  must  be  provided 
whereby  an  audible  tone  is  heard  while  the 
unmodulated  carrier  is  being  received,  the  tone 
stopping  immediately  when  the  carrier  is  in- 
terrupted. 

The  most  simple  means  of  accomplishing 
this  is  to  feed  a locally  generated  carrier  of 
a slightly  different  frequency  into  the  same 
detector,  so  that  the  incoming  signal  will  mix 
with  it  to  form  an  audible  heat  note.  The  dif- 
ference frequency,  or  heterodyne  as  the  beat 
note  is  known,  will  of  course  stop  and  start 
in  accordance  with  the  incoming  c-w  radio- 
telegraph signal,  because  the  audible  hetero- 
dyne can  exist  only  when  both  the  incoming 
and  the  locally  generated  carriers  are  present. 


The  Autodyne  The  local  signal  which  is  used 
Detector  to  beat  with  the  desired  c-w 

signal  in  the  detector  may  be 
supplied  by  a separate  low-power  oscillator 
in  the  receiver  itself,  or  the  detector  may  be 
made  to  self-oscillate,  and  thus  serve  the 
dual  purpose  of  detector  and  oscillator.  A de- 
tector which  self-oscillates  to  provide  a beat 
note  is  known  as  an  autodyne  detector,  and 
the  process  of  obtaining  feedback  between  the 
detector  plate  and  grid  is  called  re  generation. 

An  autodyne  detector  is  most  sensitive  when 
it  is  barely  oscillating,  and  for  this  reason 
a regeneration  control  is  always  included  in 
the  circuit  to  adjust  the  feedback  to  the  proper 
amount.  The  regeneration  control  may  be  either 
a variable  capacitor  or  a variable  resistor, 
as  shown  in  figure  2. 

With  the  detector  regenerative  but  not  os- 
cillating, it  is  also  quite  sensitive.  When  the 
circuit  is  adjusted  to  operate  in  this  manner, 
modulated  signals  may  be  received  with  con- 
siderably greater  strength  than  with  a non- 
regenerative  detector. 


HANDBOOK 


S u p e r r e g e n e r a t i V e Detectors  209 


12-2  Superregenerative 

Receivers 

At  ultra-high  frequencies,  when  it  is  de- 
sired to  keep  weight  and  cost  at  a minimum, 
a special  form  of  the  regenerative  receiver 
known  as  the  superregenerator  is  often  used 
for  radiotelephony  reception.  The  superregen- 
erator is  essentially  a regenerative  receiver 
with  a means  provided  to  throw  the  detector 
rapidly  in  and  out  of  oscillation.  The  frequency 
at  which  the  detector  is  made  to  go  in  and  out 
of  oscillation  varies  with  the  frequency  to  be 
received,  but  is  usually  between  20,000  and 
500,000  times  a second.  This  superregenera- 
tive action  considerably  increases  the  sen- 
sitivity of  the  oscillating  detector  so  that  the 
usual  ‘^background  hiss**  is  greatly  amplified 
when  no  signal  is  being  received.  This  hiss 
diminishes  in  proportion  to  the  strength  of  the 
received  signal,  loud  signals  eliminating  the 
hiss  entirely. 

Quench  There  are  two  systems  in  common 
Methods  use  for  causing  the  detector  to  break 
in  and  out  of  oscillation  rapidly.  In 
one,  a separate  interruption-frequency  oscilla- 
tor is  arranged  so  as  to  vary  the  voltage  rapid- 
ly on  one  of  the  detector  tube  elements  (usual- 
ly the  plate,  sometimes  the  screen)  at  the  high 
rate  necessary.  The  interruption-frequency 
oscillator  commonly  uses  a conventional  tick- 
ler-feedback circuit  with  coils  appropriate  for 
its  operating  frequency. 

The  second,  and  simplest,  type  of  super- 
regenerative  detector  circuit  is  arranged  so 
as  to  produce  its  own  interruption  frequency 
oscillation,  without  the  aid  of  a separate  tube. 
The  detector  tube  damps  (or  “quenches**)  itself 
out  of  signal-frequency  oscillation  at  a high 
rate  by  virtue  of  the  use  of  a high  value  of 
grid  leak  and  proper  size  plate-blocking  and 
grid  capacitors,  in  conjunction  with  an  excess 
of  feedback.  In  this  type  of  “self-quenched** 
detector,  the  grid  leak  is  quite  often  returned 
to  the  positive  side  of  the  power  supply  (through 
the  coil)  rather  than  to  the  cathode.  A repre- 
sentative self-quenched  superregenerative  de- 
tector circuit  is  shown  in  figure  3. 

Except  where  it  is  impossible  to  secure 
sufficient  regenerative  feedback  to  permit 
superregeneration,  the  self-quenching  circuit 
is  to  be  preferred;  it  is  simpler,  is  self-adjust- 
ing as  regards  quenching  amplitude,  and  can 
have  good  quenching  wave  form.  To  obtain 
as  good  results  with  a separately  quenched 
superregenerator,  very  careful  design  is  re- 
quired. However,  separately  quenched  circuits 
are  useful  when  it  is  possible  to  make  a cer- 
tain tube  oscillate  on  a very  high  frequency 
but  it  is  impossible  to  obtain  enough  regenera- 
tion for  self-quenching  action. 


Figure  3 

SUPERREGENERATIVE  DETECTOR  CIRCUIT 

A se/f-quenchec/  superregenerof/ve  detector  such 
as  illustrated  above  is  capable  of  giving  good 
sensitivity  in  the  v-h-f  range.  However,  the  circuit 
has  the  disadvantage  that  its  selectivity  is  rela- 
tively poor.  Also,  such  a circuit  should  be  pre- 
ceded by  an  r-f  stage  to  suppress  the  radiation  of 
a signal  by  the  oscillating  detector. 


The  optimum  quenching  frequency  is  a func- 
tion of  the  signal  frequency.  As  the  operating 
frequency  goes  up,  so  does  the  optimum  quench- 
ing frequency.  When  the  quench  frequency  is 
too  low,  maximum  sensitivity  is  not  obtained. 
When  it  is  too  high,  both  sensitivity  and  selec- 
tivity suffer.  In  fact,  the  optimum  quench  fre- 
quency for  an  operating  frequency  below  15  Me. 
is  in  the  audible  range.  This  makes  the  super- 
regenerator  impracticable  for  use  on  the  lower 
frequencies. 

The  high  background  noise  or  hiss  which 
is  heard  on  a properly  designed  superregener- 
ator when  no  signal  is  being  received  is  not 
the  quench  frequency  component;  it  is  tube 
and  tuned  circuit  fluctuation  noise,  indicating 
that  the  receiver  is  extremely  sensitive. 

A moderately  strong  signal  will  cause  the 
background  noise  to  disappear  completely, 
because  the  superregenerator  has  an  inherent 
and  instantaneous  automatic  volume  control 
characteristic.  This  same  a-v-c  characteristic 
makes  the  receiver  comparatively  insensitive 
to  impulse  noise  such  as  ignition  pulses— a 
highly  desirable  feature.  This  characteristic 
also  results  in  appreciable  distortion  of  a re- 
ceived radiotelephone  signal,  but  not  enough 
to  affect  the  intelligibility. 

The  selectivity  of  a superregenerator  is 
rather  poor  as  compared  to  a superheterodyne, 
but  is  surprisingly  good  for  so  simple  a re- 
ceiver when  figured  on  a percentage  basis 
rather  than  absolute  kc.  bandwidth. 

FM  Reception  A'  superregenerative  receiver 
will  receive  frequency  modu- 
lated signals  with  results  comparing  favorably 
with  amplitude  modulation  if  the  frequency 
swing  of  the  FM  transmitter  is  sufficiently 
high.  For  such  reception,  the  receiver  is  de- 
tuned slightly  to  either  side  of  resonance. 


210  Radio  Receiver  Fundamentals 


THE  RADIO 


^I2AT7 


THE  FREMODYNE  SUPERREGENERATIVE 
SUPERHETERODYNE  DETECTOR  FOR 
FREQUENCY  MODULATED  SIGNALS 


Superregenerative  receivers  radiate  a strong, 
broad,  and  rough  signal.  For  this  reason,  it  is 
necessary  in  most  applications  to  employ  a 
radio  frequency  amplifier  stage  ahead  of  the 
detector,  with  thorough  shielding  throughout 
the  receiver. 

The  Fremodyne  The  H a z e 1 1 i n e - Fremodyne 
Detector  superregenerative  circuit  is 

expressly  designed  for  re- 
ception of  FM  signals.  This  versatile  circuit 
combines  the  action  of  the  superregenerative 
receiver  with  the  supethetrodyne,  converting 
FM  signals  directly  into  audio  signals  in  one 
double  triode  tube  (figure  4).  One  section  of 
the  triode  serves  as  a superregenerative  mixer, 
producing  an  i-f  of  22  Me.,  an  i-f  amplifier,  and 
a FM  detector.  The  detector  action  is  accom- 
plished by  slope  detection  tuning  on  the  side 
of  the  i-f  selectivity  curve. 

This  circuit  greatly  reduces  the  radiated 
signal,  characteristic  of  the  superregenerative 
detector,  yet  provides  many  of  the  desirable 
features  of  the  superregenerator.  The  pass- 
band  of  the  Fremodyne  detector  is  about 
400  kc. 


12-3  Superheterodyne 

Receivers 


Because  of  its  superiority  and  nearly  unir 
versal  use  in  all  fields  of  radio  reception,  the 


Figure  5 

ESSENTIAL  UNITS  OF  A 
SUPERHETERODYNE  RECEIVER 
The  basic  portions  of  the  receiver  are  shown 
In  solid  blocks.  Practicable  receivers  em- 
ploy the  dotted  blocks  and  also  usually  in- 
clude such  additional  circuits  as  a noise 
limiter,  an  a-v-c  circuit,  and  a crystal  filter 
in  the  i-f  amplifier. 


theory  of  operation  of  the  superheterodyne 
should  be  familiar  to  every  radio  student  and 
experimenter.  The  following  discussion  con- 
cerns superheterodynes  for  amplitude-modula- 
tion reception.  It  is,  however,  applicable  in 
part  to  receivers  for  frequency  modulation. 

Principle  of  In  the  superheterodyne,  the  in- 
Operation  coming  signal  is  applied  to  a 
mixer  consisting  of  a non-linear 
impedance  such  as  a vacuum  tube  or  a diode. 
The  signal  is  mixed  with  a steady  signal  gen- 
erated locally  in  an  oscillator  stage,  with  the 
result  that  a signal  bearing  all  the  modulation 
applied  to  the  original  signal  hut  of  a fre- 
quency equal  to  the  difference  between  the 
local  oscillator  and  incoming  signal  frequen- 
cies appears  in  the  mixer  output  circuit.  The 
output  from  the  mixer  stage  is  fed  into  a fixed- 
tuned  intermediate-frequency  amplifier,  where 
it  is  amplified  and  detected  in  the  usual  man- 
ner, and  passed  on  to  the  audio  amplifier.  Fig- 
ure 5 shows  a block  diagram  of  the  fundamen- 
tal superheterodyne  arrangement.  The  basic 
components  are  shown  in  heavy  lines,  the 
simplest  superheterodyne  consisting  simply 
of  these  three  units.  However,  a good  com- 
munications receiver  will  comprise  all  of  the 
elements  shown,  both  heavy  and  dotted  blocks. 

Superheterodyne  The  advantages  of  super- 
Advantoges  heterodyne  reception  are 

directly  attributable  to  the 
use  of  the  fixed- tuned  intermediate- frequency 
(i-f)  amplifier.  Since  all  signals  are  converted 
to  the  intermediate  frequency,  this  section  of 
the  receiver  may  be  designed  for  optimum  se- 
lectivity and  high  amplification.  High  ampli- 
fication is  easily  obtained  in  the  intermediate- 
frequency  amplifier,  since  it  operates  at  a 


HANDBOOK 


The  Superhetrody ne  211 


VARIABLE-JU 


Figure  6 

TYPICAL  l-F  AMPLIFIER  STAGE 


relatively  low  frequency,  where  conventional 
pentode-type  tubes  give  adequate  voltage  gain. 
A typical  i-f  amplifier  is  shown  in  figure  6. 

From  the  diagram  it  may  be  seen  that  both 
the  grid  and  plate  circuits  ate  tuned.  The  tuned 
circuits  used  for  coupling  between  i-f  stages 
are  known  as  i-j  transformers.  These  will  be 
more  fully  discussed  later  in  this  chapter. 

Choice  of  Inter-  The  choice  of  a frequency 
mediate  Frequency  for  the  i-f  amplifier  in- 
volves several  considera- 
tions. One  of  these  considerations  concerns 
selectivity;  the  lower  the  intermediate  fre- 
quency the  greater  the  obtainable  selectivity. 
On  the  other  hand,  a rather  high  intermediate 
frequency  is  desirable  from  the  standpoint  of 
image  elimination,  and  also  for  the  reception 
of  signals  from  television  and  FM  transmitters 
and  modulated  self-controlled  oscillators,  all 
of  which  occupy  a rather  wide  band  of  frequen- 
cies, making  a broad  selectivity  characteristic 
desirable.  Images  are  a pecularity  common  to 
all  superheterodyne  receivers,  and  for  this 
reason  they  are  given  a detailed  discussion 
later  in  this  chapter. 

While  intermediate  frequencies  as  low  as 
50  kc.  are  used  where  extreme  selectivity  is 
a requirement,  and  frequencies  of  60  Me.  and 
above  are  used  in  some  specialized  forms  of 
receivers,  most  present-day  communications 
superheterodynes  use  intermediare  frequencies 
around  either  455  kc.  or  1600  kc. 

Home-type  broadcast  receivers  almost  al- 
ways use  an  i-f  in  the  vicinity  of  455  kc., 
while  auto  receivers  usually  use  a frequency 
of  about  262  kc.  The  standard  frequency  for 
the  i-f  channel  of  FM  receivers  is  10.7  Me. 
Television  receivers  use  an  i-f  which  covers 
rhe  band  between  about  21.5  and  27  Me.,  al- 
though a new  band  between  41  and  46  Me.  is 
coming  into  mote  common  usage. 

Arithmetical  Aside  from  allowing  the  use  of 

Selectivity  fixed-tuned  band-pass  amplifier 

stages,  the  superheterodyne  has 


an  overwhelming  advantage  over  the  tuned 
radio  frequency  (t-r-f)  type  of  receiver  because 
of  what  is  commonly  known  as  arithmetical 
selectivity. 

This  can  best  be  illustrated  by  considering 
two  receivers,  one  of  the  t-r-f  type  and  one  of 
the  superheterodyne  type,  both  attempting  to 
receive  a desired  signal  at  10,000  kc.  and 
eliminate  a strong  interfering  signal  at  10,010 
kc.  In  the  t-t-f  receiver,  separating  these  two 
signals  in  the  tuning  circuits  is  practically 
impossible,  since  they  differ  in  frequency  by 
only  0.1  per  cent.  However,  in  a superhetero- 
dyne with  an  intermediate  frequency  of,  for  ex- 
ample, 1000  kc.,  the  desired  signal  will  be 
converted  to  a frequency  of  1000  kc.  and  the 
interfering  signal  will  be  converted  to  a fre- 
quency of  1010  kc.,  both  signals  appearing  at 
the  input  of  the  i-f  amplifier.  In  this  case,  the 
two  signals  may  be  separated  much  mote  read- 
ily, since  they  differ  by  1 per  cent,  or  10  times 
as  much  as  in  the  first  case. 

The  Converter  The  converter  stage,  or  mixer, 
Stoge  of  a superheterodyne  receiver 

can  be  either  one  of  two  types: 
(1)  it  may  use  a single  envelope  converter 
tube,  such  as  a 6K8,  6SA7,  or  6BE6,  or  (2)  it 
may  use  two  tubes,  or  two  sets  of  elements  in 
the  same  envelope,  in  an  oscillator-mixer  ar- 
rangement. Figure  7 shows  a group  of  circuits 
of  both  types  to  illustrate  present  practice 
with  regard  to  types  of  converter  stages. 

Converter  tube  combinations  such  as  shown 
in  figures  7A  and  7B  are  relatively  simple  and 
inexpensive,  and  they  do  an  adequate  job  for 
most  applications.  With  a converter  tube  such 
as  the  6SB7-Y  or  the  6BA7  quite  satisfactory 
performance  may  be  obtained  for  the  reception 
of  relatively  strong  signals  (as  for  example 
FM  broadcast  reception)  up  to  frequencies  in 
excess  of  100  Me.  However,  rhe  equivalent  in- 
put noise  resistance  of  such  tubes  is  of  the 
order  of  200,000  ohms,  which  is  a rather  high 
value  indeed.  So  such  tubes  are  not  suited  for 
operation  without  an  r-f  stage  in  the  high- 
frequency  range  if  weak-signal  reception  is 
desired. 

The  6L7  mixer  circuit  shown  in  figure  7C, 
and  rhe  6BA7  circuit  of  figure  7D,  also  are 
characterized  by  an  equivalent  input  noise  re- 
sistance of  several  hundred  thousand  ohms,  so 
that  these  also  must  be  preceded  by  one  or 
mote  r-f  stages  with  a fairly  high  gain  pet 
stage  if  a low  noise  factor  is  desired  of  the 
complete  receiver. 

However,  the  circuit  arrangements  shown 
at  figures  7F  and  6F  are  capable  of  low-noise 
operation  within  themselves,  so  that  these 
circuits  may  be  fed  directly  from  the  antenna 
without  an  t-f  stage  and  still  provide  a good 
noise  factor  to  the  complete  receiver.  Note 


212  Radio  Receiver  Fundamentals 


THE  RADIO 


that  both  these  circuits  use  control-grid  in- 
jection of  both  the  incoming  signal  and  the 
local-oscillator  voltage.  Hence,  paradoxically, 
circuits  such  as  these  should  be  preceded  by 
an  r-f  stage  if  local- oscillator  radiation  is  to 
be  held  to  any  reasonable  value  of  field  in- 
tensity. 


quency  limit  for  conventional  mixer  stages, 
mixers  of  the  diode  type  are  most  commonly 
employed.  The  diode  may  be  either  a vacuum- 
tube  heater  diode  of  a special  u-h-f  design 
such  as  the  9005,  or  it  may  be  a crystal  diode 
of  the  general  type  of  the  1N21  through  1N28 
series. 


Diode  Mixers  As  the  frequency  of  operation  of 
a superheterodyne  receiver  is  in- 
creased above  a few  hundred  megacycles  the 
signal-to-noise  ratio  appearing  in  the  plate 
circuit  of  the  mixer  tube  when  triodes  or  pen- 
todes are  employed  drops  to  a prohibitively 
low  value.  At  frequencies  above  the  upper-fte- 


12-4  Mixer  Noise 

and  Images 

The  effects  of  mixer  noise  and  images  are 
troubles  common  to  all  superheterodynes.  Since 


HANDBOOK 


Mixer  Characteristics  213 


both  these  efiects  can  largely  be  obviated  by 
the  same  remedy,  they  will  be  considered  to- 
gether. 

Mixer  Noise  Mixer  noise  of  the  shot-effect 
type,  which  is  evidenced  by  a 
hiss  in  the  audio  output  of  the  receiver,  is 
caused  by  small  irregularities  in  the  plate  cur- 
rent in  the  mixer  stage  and  will  mask  weak 
signals.  Noise  of  an  identical  nature  is  gen- 
erated in  an  amplifier  stage,  but  due  to  the 
fact  that  the  conductance  in  the  mixer  stage 
is  considerably  lower  than  in  an  amplifier 
stage  using  the  same  tube,  the  proportion  of 
inherent  noise  present  in  a mixer  usually  is 
considerably  greater  than  in  an  amplifier  stage 
using  a comparable  tube. 

Although  this  noise  cannot  be  eliminated, 
its  effects  can  be  greatly  minimized  by  plac- 
ing sufficient  signal-frequency  amplification 
having  a high  signal-to-noise  ratio  ahead  of 
the  mixer.  This  remedy  causes  the  signal  out- 
put from  the  mixer  to  be  large  in  proportion  to 
the  noise  generated  in  the  mixer  stage.  In- 
creasing the  gain  after  the  mixer  will  be  of  no 
advantage  in  eliminating  mixer  noise  difficul- 
ties; greater  selectivity  after  the  mixer  will 
help  to  a certain  extent,  but  cannot  be  carried 
too  far,  since  this  type  of  selectivity  decreases 
the  i-f  band-pass  and  if  carried  too  far  will 
not  pass  the  sidebands  that  are  an  essential 
part  of  a voice-modulated  signal. 

Triode  Mixers  A ttiode  having  a high  trans- 
conductance  is  the  quietest 
mixer  tube,  exhibiting  somewhat  less  gain  but 
a better  signal-to-noise  ratio  than  a compar- 
able multi-grid  mixer  tube.  However,  below  30 
Me.  it  is  possible  to  construct  a receiver  that 
will  get  down  to  the  atmospheric  noise  level 
without  resorting  to  a ttiode  mixer.  The  addi- 
tional difficulties  experienced  in  avoiding 
pulling,  undesirable  feedback,  etc.,  when  using 
a triode  with  control-grid  injection  tend  to  make 
multi-grid  tubes  the  popular  choice  for  this 
application  on  the  lower  frequencies. 

On  very  high  frequencies,  where  set  noise 
rather  than  atmospheric  noise  limits  the  weak 
signal  response,  ttiode  mixers  ate  more  widely 
used.  A 6] 6 miniature  twin  ttiode  with  grids 
in  push-pull  and  plates  in  parallel  makes  an 
excellent  mixer  up  to  about  600  Me. 

Injection  The  amplitude  of  the  injection  volt- 
Voltoge  age  will  affect  the  conversion  trans- 
conductance of  the  mixer,  and  there- 
fore should  be  made  optimum  if  maximum  sig- 
nal-to-noise ratio  is  desired.  If  fixed  bias  is 
employed  on  the  injection  grid,  the  optimum 
injection  voltage  is  quite  critical.  If  cathode 
bias  is  used,  the  optimum  voltage  is  not  so 
critical;  and  if  grid  leak  bias  is  employed,  the 


optimum  injection  voltage  is  not  at  all  critical 
just  so  it  is  adequate.  Typical  optimum  in- 
jection voltages  will  run  from  1 to  10  volts  for 
control  grid  injection,  and  45  volts  or  so  for 
screen  or  suppressor  grid  injection. 

Images  There  always  are  two  signal  frequen- 
cies which  will  combine  with  a given 
frequency  to  produce  the  same  difference  fre- 
quency. For  example:  assume  a superhetero- 
dyne with  its  oscillator  operating  on  a higher 
frequency  than  the  signal,  which  is  common 
practice  in  present  superheterodynes,  tuned  to 
receive  a signal  at  14,100  kc.  Assuming  an 
i-f  amplifier  frequency  of  450  kc.,  the  mixer 
input  circuit  will  be  tuned  to  14,100  kc.,  and 
the  oscillator  to  14,100  plus  450,  or  14,550  kc. 
Now,  a strong  signal  at  the  oscillator  frequen- 
cy plus  the  intermediate  frequency  (14,550 
plus  450,  or  15,000  kc.)  will  also  give  a dif- 
ference frequency  of  450  kc.  in  the  mixer  out- 
put and  will  be  heard  also.  Note  that  the  image 
is  always  twice  the  intermediate  frequency 
away  from  the  desired  signal.  Images  cause 
repeat  points  on  the  tuning  dial. 

The  only  way  that  the  image  could  be  elimi- 
nated in  this  particular  case  would  be  to  make 
the  selectivity  of  the  mixer  input  circuit,  and 
any  circuits  preceding  it,  great  enough  so  that 
the  15,000-kc.  signal  never  reaches  the  mixer 
grid  in  sufficient  amplitude  to  produce  inter- 
ference. 

For  any  particular  intermediate  frequency, 
image  interference  troubles  become  increas- 
’"gly  greater  as  the  frequency  to  which  the 
signal- frequency  portion  of  the  receiver  is 
tuned  is  increased.  This  is  due  to  the  fact  that 
the  percentage  difference  between  the  desired 
frequency  and  the  image  frequency  decreases 
as  the  receiver  is  tuned  to  a higher  frequency. 
The  ratio  of  strength  between  a signal  at  the 
image  frequency  and  a signal  at  the  frequency 
to  which  the  receiver  is  tuned  producing  equi 
output  is  known  as  the  image  ratio.  The  higher 
this  ratio,  the  better  the  receiver  in  regard  to 
image-interference  troubles. 

With  but  a single  tuned  circuit  between  the 
mixer  grid  and  the  antenna,  and  with  400-500 
kc.  i-f  amplifiers,  image  ratios  of  60  db  and 
over  ate  easily  obtainable  up  to  frequencies 
around  2000  kc.  Above  this  frequency,  greater 
selectivity  in  the  mixer  grid  circuit  through 
the  use  of  additional  tuned  circuits  between 
the  mixer  and  the  antenna  is  necessary  if  a 
good  image  ratio  is  to  be  maintained. 

12-5  R-F  Stages 

Since  the  necessary  tuned  circuits  between 
the  mixer  and  the  antenna  can  be  combined 
with  tubes  to  form  r-f  amplifier  stages,  the 


214  Radio  Receiver  Fundamentals 


THE  RADIO 


PENTODE 


Figure  8 

TYPICAL  PENTODE  R-F  AMPLIFIER  STAGE 


reduction  of  the  effects  of  mixer  noise  and  the 
increasing  of  the  image  ratio  can  be  accom- 
plished in  a single  section  of  the  receiver. 
When  incorporated  in  the  receiver,  this  sec- 
tion is  known  simply  as  an  r-/  amplifier;  when 
it  is  a separate  unit  with  a separate  tuning 
control  it  is  often  known  as  a preselector. 
Either  one  or  two  stages  are  commonly  used 
in  the  preselector  or  r-f  amplifier.  Some  pre- 
selectors use  regeneration  to  obtain  still 
greater  amplification  and  selectivity.  An  r-f 
amplifier  or  preselector  embodying  more  than 
two  stages  rarely  ever  is  employed  since  two 
stages  will  ordinarily  give  adequate  gain  to 
override  mixer  noise. 

R-F  Stages  In  Generally  speaking,  atmos- 
the  V-H-F  Range  pheric  noise  in  the  frequency 
range  above  30  Me.  is  quite 
low— so  low,  in  fact,  that  the  noise  generated 
within  the  receiver  itself  is  greater  than  the 
noise  received  on  the  antenna.  Hence  it  is  of 
the  greatest  importance  that  internally  gener- 
ated noise  be  held  to  a minimum  in  a receiver. 
At  frequencies  much  above  300  Me.  there  is 
not  too  much  that  can  be  done  at  the  present 
state  of  the  art  in  the  direction  of  reducing 
receiver  noise  below  that  generated  in  the  con- 
verter stage.  But  in  the  v-h-f  range,  between 
30  and  300  Me.,  the  receiver  noise  factor  in  a 
well  designed  unit  is  determined  by  the  char- 
acteristics of  the  first  r-f  stage. 

The  usual  v-h-f  receiver,  whether  for  com- 
munications or  for  FM  or  TV  reception,  uses 
a miniature  pentode  for  the  first  r-f  amplifier 
stage.  The  6AK5  is  the  best  of  presently  avail- 
able types,  with  the  6CB6  and  the  6DC6  close- 
ly approaching  the  6AK5  in  performance.  But 
when  gain  in  the  first  r-f  stage  is  not  so  im- 
portant, and  the  best  noise  factor  must  be  ob- 
tained, the  first  r-f  stage  usually  uses  a triode. 

Shown  in  figure  9 are  four  commonly  used 
types  of  triode  r-f  stages  for  use  in  the  v-h-f 
range.  The  circuit  at  (A)  uses  few  components 
and  gives  a moderate  amount  of  gain  with  very 
low  noise.  It  is  most  satisfactory  when  the 
first  r-f  stage  is  to  be  fed  directly  from  a low- 


6AB4,  6J6, 


CATHODE-COUPLED  6J6 


© 


LOW  NOISE  Ln 


Ln 


Figure  9 

TYPICAL  TRIODE  V-H-F 
R-F  AMPLIFIER  STAGES 
Triode  r-f  stages  contribute  the  least  amount  of 
noise  output  for  a given  signal  level,  hence  their 
frequent  use  in  the  v-h-f  range. 


impedance  coaxial  transmission  line.  Figure 
9 (B)  gives  somewhat  more  gain  than  (A),  but 
requires  an  input  matching  circuit.  The  effec- 
tive gain  of  this  circuit  is  somewhat  reduced 
when  it  is  being  used  to  amplify  a broad  band 
of  frequencies  since  the  effective  of  the 
cathode- coupled  dual  tube  is  somewhat  less 


HANDBOOK 


The  Cascode  Amplifier  215 


Figure  10 

TYPICAL  DOUBLE-CONVERSION  SUPERHETERODYNE  RECEIVERS 
Illustrated  at  (A)  is  the  basic  circuit  of  a commercio/  double- conversion  superheterodyne  receiver.  At  (B)  is 
illustrated  the  application  of  an  accessory  sharp  i-f  channel  for  obtaining  improved  selectivity  from  a con- 
ventional communications  receiver  through  the  use  of  the  double-conversion  principle. 


than  half  the  of  either  of  the  two  tubes 
taken  alone. 

The  Cascode  The  Cascode  r-f  amplifier,  de- 
Amplifier  veloped  at  the  MIT  Radiation 

Laboratory  during  World  War  II, 
is  a low  noise  circuit  employing  a grounded 
cathode  triode  driving  a grounded  grid  triode, 
as  shown  in  figure  9C.  The  stage  gain  of  such 
a circuit  is  about  equal  to  that  of  a pentode 
tube,  while  the  noise  figure  remains  at  the  low 
level  of  a triode  tube.  Neutralization  of  the 
first  triode  tube  is  usually  unnecessary  below 
50  Me.  Above  this  frequency,  a definite  im- 
provement in  the  noise  figure  may  be  obtained 
through  the  use  of  neutralization.  The  neutral- 
izing coil,  Ln,  should  resonate  at  the  operat- 
ing frequency  with  the  gtid-plate  capacity  of 
the  first  triode  tube. 

The  6BQ7A  and  6BZ7  tubes  ate  designed  for 
use  in  cascode  circuits,  and  may  be  used  to 
good  advantage  in  the  144  Me.  and  220  Me.  am- 
ateur bands  (figure  9D).  For  operation  at  higher 
frequencies,  the  6AJ4  tube  is  recommended. 

Double  Conversion  As  previously  mentioned, 
the  use  of  a higher  inter- 
mediate frequency  will  also  improve  the  image 


ratio,  at  the  expense  of  i-f  selectivity,  by 
placing  the  desired  signal  and  the  image  far- 
ther apart.  To  give  both  good  image  ratio  at 
the  higher  frequencies  and  good  selectivity  in 
the  i-f  amplifier,  a system  known  as  double 
conversion  is  sometimes  employed.  In  this  sys- 
tem, the  incoming  signal  is  first  converted  to 
a rather  high  intermediate  frequency,  and  then 
amplified  and  again  converted,  this  time  to  a 
much  lower  frequency.  The  first  intermediate 
frequency  supplies  the  necessary  wide  separa- 
tion between  the  image  and  the  desired  sig- 
nal, while  the  second  one  supplies  the  bulk  of 
the  i-f  selectivity. 

The  double-conversion  system,  as  illus- 
trated in  figure  10,  is  receiving  two  general 
types  of  application  at  the  present  time.  The 
first  application  is  for  the  purpose  of  attaining 
extremely  good  stability  in  a communications 
receiver  through  the  use  of  crystal  control  of 
the  first  oscillator.  In  such  an  arrangement, 
as  used  in  several  types  of  Collins  receivers, 
the  first  oscillator  is  crystal  controlled  and  is 
followed  by  a tunable  i-f  amplifier  which  then 
is  followed  by  a mixer  stage  and  a fixed-tuned 
i-f  amplifier  on  a much  lower  frequency. 
Through  such  a circuit  arrangement  the  sta- 
bility of  the  complete  receiver  is  equal  to  the 


216  Radio  Receiver  Fundamentals 


THE  RADIO 


stability  of  the  oscillator  which  feeds  the  sec- 
ond mixer,  while  the  selectivity  is  determined 
by  the  bandwidth  of  the  second,  fixed  i-f  am- 
plifier. 

The  second  common  application  of  the 
double-conversion  principle  is  for  the  purpose 
of  obtaining  a very  high  degree  of  selectivity 
in  the  complete  communications  receiver.  In 
this  type  of  application,  as  illustrated  in  fig- 
ure 10  (B),  a conventional  communications  re- 
ceiver is  modified  in  such  a manner  that  its 
normal  i-f  amplifier  (which  usually  is  in  the 
450  to  915  kc.  range)  instead  of  being  fed  to 
a demodulator  and  then  to  the  audio  system, 
is  alternatively  fed  to  a fixed-tune  mixer  stage 
and  then  into  a much  lower  intermediate  fre- 
quency amplifier  before  the  signal  is  demodu- 
lated and  fed  to  the  audio  system.  The  acces- 
sory i-f  amplifier  system  (sometimes  called  a 
Q5’er)  normally  is  operated  on  a frequency  of 
175  kc.,  85  kc.,  or  50  kc. 


12-6  Signal-Frequency 

Tuned  Circuits 

The  signal-frequency  tuned  circuits  in  high- 
frequency  superheterodynes  and  tuned  radio 
frequency  types  of  receivers  consist  of  coils 
of  either  the  solenoid  or  universal- wound  types 
shunted  by  variable  capacitors.  It  is  in  these 
tuned  circuits  that  the  causes  of  success  or 
failure  of  a receiver  often  lie.  The  universal- 
wound  type  coils  usually  are  used  at  frequen- 
cies below  2000  kc.;  above  this  frequency  the 
single-layer  solenoid  type  of  coil  is  more 
satisfactory. 

Impedance  The  two  factors  of  greatest  signi- 
and  Q ficance  in  determining  the  gain- 

per-stage  and  selectivity,  respec- 
tively, of  a tuned  amplifier  are  tuned-circuit 
impedance  and  tuned-circuit  Q.  Since  the  re- 
sistance of  modern  capacitors  is  low  at  ordi- 
nary frequencies,  the  resistance  usually  can 
be  considered  to  be  concentrated  in  the  coil. 
The  resistance  to  be  considered  in  making  Q 
determinations  is  the  r-f  resistance,  not  the 
d-c  resistance  of  the  wire  in  the  coil.  The  lat- 
ter ordinarily  is  low  enough  that  it  may  be 
neglected.  The  increase  in  r-f  resistance  over 
d-c  resistance  primarily  is  due  to  skin  effect 
and  is  influenced  by  such  factors  as  wire  size 
and  type,  and  the  proximity  of  metallic  objects 
or  poor  insulators,  such  as  coil  forms  with 
high  losses.  Higher  values  of  Q lead  to  better 
selectivity  and  increased  r-f  voltage  across 
the  tuned  circuit.  The  increase  in  voltage  is 
due  to  an  increase  in  the  circuit  impedance 
with  the  higher  values  of  Q. 


TOA.V.C.  ■=■  +B 


Figure  1 1 

ILLUSTRATING  "COMMON  POINT” 

by-passing 

To  reduce  the  detrimental  effects  of  cathode  cir- 
cuit inductance  in  v-h-f  stages,  all  by-pass  ca- 
pacitors should  be  returned  to  the  cathode  terminal 
at  the  socket.  Tubes  with  two  cathode  leads  can 
give  improved  performance  if  the  grid  return  is 
made  to  one  cathode  terminal  while  the  plate  and 
screen  by-pass  returns  are  made  ta  the  cathode 
terminal  which  is  connected  to  the  suppressor 
within  the  tube. 


Frequently  it  is  possible  to  secure  an  in- 
crease in  impedance  in  a resonant  circuit,  and 
consequently  an  increase  in  gain  from  an  am- 
plifier stage,  by  increasing  the  reactance 
through  the  use  of  larger  coils  and  smaller 
tuning  capacitors  (higher  L/C  ratio). 

Input  Resistance  Another  factor  which  in- 
fluences the  operation  of 
tuned  circuits  is  the  input  resistance  of  the 
tubes  placed  across  these  circuits.  At  broad- 
cast frequencies,  the  input  resistance  of  most 
conventional  r-f  amplifier  tubes  is  high  enough 
so  that  it  is  not  bothersome.  But  as  the  fre- 
quency is  increased,  the  input  resistance  be- 
comes lower  and  lower,  until  it  ultimately 
reaches  a value  so  low  that  no  amplification 
can  be  obtained  from  the  r-f  stage. 

The  two  contributing  factors  to  the  decrease 
in  input  resistance  with  increasing  frequency 
are  the  transit  time  required  by  an  electron 
traveling  between  the  cathode  and  grid,  and 
the  inductance  of  the  cathode  lead  common 
to  both  the  plate  and  grid  circuits.  As  the 
frequency  becomes  higher,  the  transit  time 
can  become  an  appreciable  portion  of  the  time 
required  by  an  r-f  cycle  of  the  signal  voltage, 
and  current  will  actually  flow  into  the  grid. 
The  result  of  this  effect  is  similar  to  that 
which  would  be  obtained  by  placing  a resis- 
tance between  the  tube’s  grid  and  cathode. 

Superheterodyne  Because  the  oscillator  in  a 

Tracking  superheterodyne  operates 

"offset”  from  the  other  front 
end  circuits,  it  is  necessary  to  make  special 
provisions  to  allow  the  oscillator  to  track 
when  similar  tuning  capacitor  sections  are 


HANDBOOK 


Tuning  Circuits  217 


Figure  12 

SERIES  TRACKING  EMPLOYED 
IN  THE  H-F  OSCILLATOR  OF  A 
SUPERHETERODYNE 

The  series  tracking  capacitor  permits  the  use  of 
identical  gangs  in  a ganged  capacitor^  since  the 
tracking  capacitor  slows  down  the  rate  of  frequen- 
cy change  in  the  oscillator  so  that  a constant  dif‘ 
ference  In  frequency  between  the  oscillator  and 
the  r-f  stage  (equal  to  the  I'-f  amplifier  frequency) 
may  be  maintained. 


ganged.  The  usual  method  of  obtaining  good 
tracking  is  to  operate  the  oscillator  on  the 
high-frequency  side  of  the  mixer  and  use  a 
series  tracking  capacitor  to  slow  down  the 
tuning  rate  of  the  oscillator.  The  oscillator 
tuning  rate  must  be  slower  because  it  covets 
a smaller  range  than  does  the  mixer  when  both 
are  expressed  as  a percentage  of  frequency. 
At  frequencies  above  7000  kc.  and  with  ordi- 
nary intermediate  frequencies,  the  difference 
in  percentage  between  the  two  tuning  ranges 
is  so  small  that  it  may  be  disregarded  in  re- 
ceivers designed  to  cover  only  a small  range, 
such  as  an  amateur  band. 

A mixer  and  oscillator  tuning  arrangement 
in  which  a series  tracking  capacitor  is  provid- 
ed is  shown  in  figure  12.  The  value  of  the 
tracking  capacitor  varies  considerably  with 
different  intermediate  frequencies  and  tuning 
ranges,  capacitances  as  low  as  .0001  pfd. 
being  used  at  the  lower  tuning-range  frequen- 
cies, and  values  up  to  .01  pfd.  being  used  at 
the  higher  frequencies. 

Superheterodyne  receivers  designed  to  cover 
only  a single  frequency  range,  such  as  the 
standard  broadcast  band,  sometimes  obtain 
tracking  between  the  oscillator  and  the  r-f  cir- 
cuits by  cutting  the  variable  plates  of  the  os- 
cillator tuning  section  to  a different  shape 
from  those  used  to  tune  the  t-f  stages. 

Frequency  Range  The  frequency  to  which  a 
Selection  receiver  responds  may  be 

varied  by  changing  the  size 
of  either  the  coils  or  the  capacitors  in  the  tun- 
ing circuits,  or  both.  In  short-wave  receivers 


Figure  13 

BANDSPREAD  CIRCUITS 
Parallel  bandspread  is  illustrated  at  (A)  and  (B), 
series  bandspread  at  fO,  and  tapperhcoil  band- 
spread  at  (D), 


a combination  of  both  methods  is  usually  em- 
ployed, the  coils  being  changed  from  one  band 
to  another,  and  variable  capacitors  being  used 
to  tune  the  receiver  across  each  band.  In  prac- 
tical receivers,  coils  may  be  changed  by  one 
of  two  methods:  a switch,  controllable  from 
the  panel,  may  be  used  to  switch  coils  of  dif- 
ferent sizes  into  the  tuning  circuits  or,  alter- 
natively, coils  of  different  sizes  may  be 
plugged  manually  into  the  receiver,  the  con- 
nection into  the  tuning  circuits  being  made  by 
suitable  plugs  on  the  coils.  Where  there  are 
several  plug-in  coils  for  each  band,  they  ate 
sometimes  arranged  to  a single  mounting  strip, 
allowing  them  all  to  be  plugged  in  simultan- 
eously. 

Bandspread  In  receivers  using  large  tuning 
Tuning  capacitors  to  covet  the  short- 

wave spectrum  with  a minimum 
of  coils,  tuning  is  likely  to  be  quite  difficult, 
owing  to  the  large  frequency  range  coveted  by 
a small  rotation  of  the  variable  capacitors. 
To  alleviate  this  condition,  some  method  of 
slowing  down  the  tuning  rate,  or  bandspread- 
ing, must  be  used. 

Quantitatively,  bandspread  is  usually  desig- 
nated as  being  inversely  proportional  to  the 
range  coveted.  Thus,  a large  amount  of  band- 
spread  indicates  that  a small  frequency  range 
is  covered  by  the  bandspread  control.  Con- 
versely, a small  amount  of  bandspread  is  taken 
to  mean  that  a large  frequency  range  is  coveted 
by  the  bandspread  dial. 

Types  of  Bandspreading  systems  are  of 
Bandspread  two  general  types:  electrical  and 

mechanical.  Mechanical  systems 
are  exemplified  by  high-ratio  dials  in  which 
the  tuning  capacitors  rotate  much  mote  slowly 


218  Radio  Receiver  Fundamentals 


THE  RADIO 


than  the  dial  knob.  In  this  system,  there  is 
often  a separate  scale  or  pointer  either  con- 
nected or  geared  to  the  dial  knob  to  facilitate 
accurate  dial  readings.  However,  there  is  a 
practical  limit  to  the  amount  of  mechanical 
bandspread  which  can  be  obtained  in  a dial 
and  capacitor  before  the  speed-reduction  unit 
and  capacitor  bearings  become  prohibitively 
expensive.  Hence,  most  receivers  employ  a 
combination  of  electrical  and  mechanical  band- 
spread.  In  such  a system,  a moderate  reduc- 
tion in  the  tuning  rate  is  obtained  in  the  dial, 
and  the  test  of  the  reduction  obtained  by  elec- 
trical bandspreading. 

Stray  Circuit  In  this  book  and  in  other  radio 
Capacitance  literature,  mention  is  sometimes 
made  of  stray  or  circuit  capaci- 
tance. This  capacitance  is  in  the  usual  sense 
defined  as  the  capacitance  remaining  across 
a coil  when  all  the  tuning,  bandspread,  and 
padding  capacitors  across  the  circuit  are  at 
their  minimum  capacitance  setting. 

Circuit  capacitance  can  be  attributed  to  two 
general  sources.  One  source  is  that  due  to  the 
input  and  output  capacitance  of  the  tube  when 
its  cathode  is  heated.  The  input  capacitance 
varies  somewhat  from  the  static  value  when 
the  tube  is  in  actual  operation.  Such  factors 
as  plate  load  impedance,  grid  bias,  and  fre- 
quency will  cause  a change  in  input  capaci- 
tance. However,  in  all  except  the  extremely 
high-transconductance  tubes,  the  published 
measured  input  capacitance  is  reasonably  close 
to  the  effective  value  when  the  tube  is  used 
within  its  recommended  frequency  range.  But 
in  the  high-transconductance  types  the  effec- 
tive capacitance  will  vary  considerably  from 
the  published  figures  as  operating  conditions 
ate  changed. 

The  second  source  of  circuit  capacitance, 
and  that  which  is  more  easily  controllable,  is 
that  contributed  by  the  minimum  capacitance 
of  the  variable  capacitors  across  the  circuit 
and  that  due  to  capacitance  between  the  wir- 
ing and  ground.  In  well-designed  high-fre- 
quency receivers,  every  effort  is  made  to  keep 
this  portion  of  the  circuit  capacitance  at  a 
minimum  since  a large  capacitance  reduces 
the  tuning  range  available  with  a given  coil 
and  prevents  a good  L/C  ratio,  and  conse- 
quently a high-impedance  tuned  circuit,  from 
being  obtained. 

A good  percentage  of  stray  circuit  capaci- 
tance is  due  also  to  distributed  capacitance 
of  the  coil  and  capacitance  between  wiring 
points  and  chassis. 

Typical  values  of  circuit  capacitance  may 
tun  from  10  to  75  ppiA.  in  high-frequency  re- 
ceivers, the  first  figure  representing  concen- 
tric-line receivers  with  acorn  or  miniature 
tubes  and  extremely  small  tuning  capacitors. 


and  the  latter  representing  all-wave  sets  with 

bandswitching,  large  tuning  capacitors,  and 
conventional  tubes. 


12-7  l-F  Tuned  Circuits 

I-f  amplifiers  usually  employ  bandpass  cir- 
cuits of  some  sort.  A bandpass  circuit  is  ex- 
actly what  the  name  implies-a  circuit  for  pass- 
ing a band  of  frequencies.  Bandpass  arrange- 
ments can  be  designed  for  almost  any  degree 
of  selectivity,  the  type  used  in  any  particular 
case  depending  upon  the  ultimate  application 
of  the  amplifier. 

1"^  Intermediate  frequency  trans- 

Transformers  formers  ordinarily  consist  of 

two  or  more  tuned  circuits  and 
some  method  of  coupling  the  tuned  circuits 
together.  Some  representative  arrangements 
are  shown  in  figure  14.  The  circuit  shown  at 
A is  the  conventional  i-f  transformer,  with  the 
coupling,  M,  between  the  tuned  circuits  being 
provided  by  inductive  coupling  from  one  coil 
to  the  other.  As  the  coupling  is  increased,  the 
selectivity  curve  becomes  less  peaked,  and 
when  a condition  known  as  critical  coupling 
is  reached,  the  top  of  the  curve  begins  to  flat- 
ten out.  When  the  coupling  is  increased  still 
more,  a dip  occurs  in  the  top  of  the  curve. 

The  windings  for  this  type  of  i-f  transformer, 
as  well  as  most  others,  nearly  always  consist 
of  small,  flat  universal-wound  pies  mounted 
either  on  a piece  of  dowel  to  provide  an  air 
core  or  on  powdered-iron  for  iron  core  i-f  trans- 
formers. The  iron-core  transformers  generally 
have  somewhat  more  gain  and  better  selectivity 
than  equivalent  air-core  units. 

The  circuits  shown  at  figure  14-B  and  C are 
quite  similar.  Their  only  difference  is  the  type 
of  mutual  coupling  used,  an  inductance  being 
used  at  B and  a capacitance  at  C.  The  opera- 
tion of  both  circuits  is  similar.  Three  reson- 
ant circuits  are  formed  by  the  components.  In 
B,  for  example,  one  resonant  circuit  is  formed 
by  Li,  Cl,  Cj  and  Lj  all  in  series.  The  fre- 
quency of  this  resonant  circuit  is  just  the  same 
as  that  of  a single  one  of  the  coils  and  capaci- 
tors, since  the  coils  and  capacitors  are  simi- 
lar in  both  sides  of  the  circuit,  and  the  reson- 
ant frequency  of  the  two  capacitors  and  the 
two  coils  all  in  series  is  the  same  as  that  of 
a single  coil  and  capacitor.  The  second  reson- 
ant frequency  of  the  complete  circuit  is  deter- 
mined by  the  characteristics  of  each  half  of 
the  circuit  containing  the  mutual  coupling  de- 
vice. In  B,  this  second  frequency  will  be  lower 
than  the  first,  since  the  resonant  frequency  of 
Li,  Cl  and  the  inductance,  M,  or  Li,  Cj  and  M 
is  lower  than  that  of  a single  coil  and  capaci- 


HANDBOOK 


l-F  Amplifiers  219 


tor,  due  to  the  inductance  of  M being  added  to 
the  circuit. 

The  opposite  effect  takes  place  at  figure 
14-C,  where  the  common  coupling  impedance 
is  a capacitor.  Thus,  at  C the  second  reson- 
ant frequency  is  higher  than  the  first.  In  either 
case,  however,  the  circuit  has  two  resonant  fre- 
quencies, resulting  in  a flat-topped  selectivity 
curve.  The  width  of  the  top  of  the  curve  is 
controlled  by  the  reactance  of  the  mutual 
coupling  component.  As  this  reactance  is  in- 
creased (inductance  made  greater,  capacitance 
made  smaller),  the  two  resonant  frequencies 
become  further  apart  and  the  cutve  is  broad- 
ened. 

In  the  circuit  of  figure  14-D,  there  is  induc- 
tive coupling  between  the  center  coil  and  each 
of  the  outer  coils.  The  result  of  this  arrange- 
ment is  that  the  center  coil  acts  as  a sharply 
tuned  coupler  between  the  other  two.  A sign^ 
somewhat  off  the  resonant  frequency  of  the 
transformer  will  not  induce  as  much  current 
in  the  center  coil  as  will  a signal  of  the  cor- 
rect frequency.  When  a smaller  current  is  in- 
duced in  the  center  coil,  it  in  turn  transfers 
a still  smaller  current  to  the  output  coil.  The 
effective  coupling  between  the  outer  coils  in- 
creases as  the  resonant  frequency  is  ap- 
proached, and  remains  nearly  constant  over  a 
small  range  and  then  decreases  again  as  the 
resonant  band  is  passed. 

Another  very  satisfactory  bandpass  arrange- 
ment, which  gives  a very  straight-sided,  flat- 
topped  cutve,  is  the  negative-mutual  arrange- 
ment shown  at  figure  14-E.  Energy  is  trans- 
ferred between  the  input  and  output  circuits  in 
this  arrangement  by  both  the  negative-mutual 
coils,  M,  and  the  common  capacitive  reactance, 
C.  The  negative-mutual  coils  are  intetwound 
on  the  same  form,  and  connected  backward. 

Transformers  usually  ate  made  tunable  over 
a small  range  to  permit  accurate  alignment  in 
the  circuit  in  which  they  ate  employed.  This 
is  accomplished  either  by  means  of  a variable 
capacitor  across  a fixed  inductance,  or  by 
means  of  a fixed  capacitor  across  a variable 
inductance.  The  former  usually  employ  either 
a mica-compression  capacitor  (designated 
"mica  tuned”),  or  a small  ait  dielectric  vari- 
able capacitor  (designated  "air  tuned”).  Those 
which  use  a fixed  capacitor  usually  employ  a 
powdered  iron  cote  on  a threaded  tod  to  vary 
the  inductance,  and  ate  known  as  "permea- 
bility tuned.” 

Shape  Factor  It  is  obvious  that  to  pass  modu- 
lation sidebands  and  to  allow 
for  slight  drifting  of  the  transmitter  carrier  fre- 
quency and  the  receiver  local  oscillator,  the 

i-f  amplifier  must  pass  not  a single  frequency 

but  a band  of  frequencies.  The  width  of  this 
pass  band,  usually  5 to  8 kc.  at  maximum 


l-F  AMPLIFIER  COUPLING 
ARRANGEMENTS 

TJe  interstage  coupling  arrangements  illustrated 
ateve  giVe  a better  shape  factor  (more  straight 
sided  selectivity  curve)  than  would  the  same  num- 
ber of  tuned  circuits  coupled  by  means  of  tubes. 


width  in  a good  communications  receiver,  is 
known  as  the  pass  band,  and  is  arbitrarily 
taken  as  the  width  between  the  two  frequen- 
cies at  which  the  response  is  attenuated  6 db, 
or  is  "6  db  down.”  However,  it  is  apparent 
that  to  discriminate  against  an  interfering  sig- 
nal which  is  stronger  than  the  desired  signal, 
much  more  than  6 db  attenuation  is  requited. 

The  attenuation  arbitrarily  taken  to  indicate 

adequate  discrimination  against  an  interfering 
signal  is  60  db. 


220  Radio  Receiver  Fundamentals 


THE  RADIO 


0 


KC. 


Figure  15 

l-F  PASS  BAND  OF  TYPICAL 
COMMUNICATIONS  RECEIVER 


It  is  apparent  that  it  is  desirable  to  have 
the  bandwidth  at  60  db  down  as  narrow  as 
possible,  but  it  must  be  done  without  making 
the  pass  band  (6  db  points)  too  narrow  for  satis- 
factory reception  of  the  desired  signal.  The 
figure  of  merit  used  to  show  the  ratio  of  band- 
width at  6 db  down  to  that  at  60  db  down  is 
designated  shape  factor.  The  ideal  i-f  curve, 
a rectangle,  would  have  a shape  factor  of  1.0. 
The  i-f  shape  factor  in  typical  communications 
receivers  runs  from  3.0  to  5.5. 

The  most  practicable  method  of  obtaining  a 
low  shape  factor  for  a given  number  of  tuned 
circuits  is  to  employ  them  in  pairs,  as  in  fig- 
ure 14-A,  adjusted  to  critical  coupling  (the 
value  at  which  two  resonance  points  just  be- 
gin to  become  apparent).  If  this  gives  too 
sharp  a "nose”  or  pass  band,  then  coils  of 
lower  Q should  be  employed,  wirh  the  coupling 
maintained  at  the  critical  value.  As  the  Q is 
lowered,  closer  coupling  will  be  required  for 
crirical  coupling. 

Conversely  if  the  pass  band  is  too  broad, 
coils  of  higher  Q should  be  employed,  the 
coupling  being  maintained  at  critical.  If  the 
pass  band  is  made  more  narrow  by  using  looser 
coupling  instead  of  raising  the  Q and  main- 
taninig  critical  coupling,  the  shape  factor  will 
not  be  as  good. 

The  pass  band  will  not  be  much  narrower 
for  several  pairs  of  identical,  critically  coupled 
tuned  circuits  than  for  a single  pair.  However, 
the  shape  factor  will  be  greatly  improved  as 
each  additional  pair  is  added,  up  to  about  5 
pairs,  beyond  which  the  improvement  for  each 
additional  pair  is  not  significant.  Commer- 
cially available  communications  receivers  of 


L C R 
I — — II—  — WWW- 


Figure  16 

ELECTRICAL  EQUIVALENT  OF 
QUARTZ  FILTER  CRYSTAL 
The  crysfal  is  equivalent  to  a very  large  value  of 
inductance  in  series  with  small  values  of  cqpacf- 
tance  and  resistance,  with  a larger  though  still 
small  value  of  capacitance  across  the  w/iofe  cir- 
cuit 'irepresenting  holder  capacitance  plus  stray 
capacitances). 


good  quality  normally  employ  3 or  4 double 
tuned  transformers  with  coupling  adjusted  to 
critical  or  slightly  less. 

The  pass  band  of  a typical  communication 
receiver  having  a 455  kc.  i-f  amplifier  is  shown 
in  figure  15. 

"Miller  As  mentioned  previously,  the  dyna- 
Effect"  mic  input  capacitance  of  a tube  varies 
slightly  with  bias.  As  a-v-c  voltage 
normally  is  applied  to  i-f  tubes  for  radiotele- 
phony reception,  the  effective  grid-cathode 
capacitance  varies  as  the  signal  strength 
varies,  which  produces  the  same  effect  as 
slight  detuning  of  the  i-f  transformer.  This 
effect  is  known  as  "Miller  effect,”  and  can 
be  minimized  to  the  extent  that  it  is  not 
troublesome  either  by  using  a fairly  low  L/C 
ratio  in  the  transformers  or  by  incorporating 
a small  amount  of  degenerative  feedback,  the 
latter  being  most  easily  accomplished  hy  leav- 
ing part  of  the  cathode  resistor  unbypassed 
for  r.f. 

Crystal  Filters  Xhe  pass  band  of  an  inter- 
mediate frequency  amplifier 
may  be  made  very  narrow  through  the  use  of  a 
piezoelectric  filter  crystal  employed  as  a 
series  resonant  circuit  in  a bridge  arrange- 
ment known  as  a crystal  filter.  The  shape  fac- 
tor is  quite  poor,  as  would  be  expected  when 
the  selectivity  is  obtained  from  the  equivalent 
of  a single  tuned  circuit,  but  the  very  narrow 
pass  band  obtainable  as  a result  of  the  ex- 
tremely high  Q of  the  crystal  makes  the  crys- 
tal filter  useful  for  c-w  telegraphy  reception. 
The  pass  hand  of  a 455  kc.  cryst^  filter  may 
be  made  as  narrow  as  50  cycles,  while  the 
narrowest  pass  band  that  can  be  obtained  with 
a 455  kc.  tuned  circuit  of  practicable  dimen- 
sions is  about  5 kc. 

The  electrical  equivalent  of  a filter  crystal 
is  shown  in  figure  16.  For  a given  frequency, 
L is  very  high,  C very  low,  and  R (assuming 


HANDBOOK 


Crystal  Filters  221 


Figure  17 

EQUIVALENT  OF  CRYSTAL 
FILTER  CIRCUIT 

For  a given  voltage  out  of  the  generator^  the  volt- 
age developed  across  Zj  depends  upon  the  ratio 
of  the  impedance  of  X to  the  sum  of  the  impedances 
of  Z and  Zi.  Because  of  the  high  Q of  the  crystal. 
Its  impedance  changes  rapidly  with  chartges  in 
frequency. 


a good  crystal  of  high  Q)  is  very  low.  Capaci- 
tance C,  represents  the  shunt  capacitance  of 
the  electrodes,  plus  the  crystal  holder  and 
wiring,  and  is  many  times  the  capacitance  of 
C.  This  makes  the  crystal  act  as  a parallel 
resonant  circuit  with  a frequency  only  slightly 
higher  than  that  of  its  frequency  of  series 
resonance.  For  crystal  filter  use  it  is  the 
series  resonant  characteristic  that  we  ace  pri- 
marily interested  in. 

The  electrical  equivalent  of  the  basic  crys- 
tal filter  circuit  is  shown  in  figure  17.  If  the 
impedance  of  Z plus  Zj  is  low  compared  to  the 
impedance  of  the  crystal  X at  resonance,  then 
the  current  flowing  through  Z,,  and  the  voltage 
developed  across  it,  will  be  almost  in  inverse 
proportion  to  the  impedance  of  X,  which  has 
a very  sharp  resonance  curve. 

If  the  impedance  of  Z plus  Zj  is  made  high 
compared  to  the  resonant  impedance  of  X,  then 
there  will  be  no  appreciable  drop  in  voltage 
across  Z,  as  the  frequency  departs  from  the 
resonant  frequency  of  X until  the  point  is 
reached  where  the  impedance  of  X approaches 
that  of  Z plus  Z,.  This  has  the  effect  of  broad- 
ening out  the  curve  of  frequency  versus  voltage 
developed  across  Zj,  which  is  another  way  of 
saying  that  the  selectivity  of  the  crystal  filter 
(but  not  the  crystal  proper)  has  been  reduced. 

In  practicable  filter  circuits  the  impedtuices 
Z and  Z,  usually  are  represented  by  some  form 
of  tuned  circuit,  but  the  basic  principle  of 
operation  is  the  same. 

Practical  Filters  It  is  necessary  to  balance 
out  the  capacitance  across 
the  crystal  holder  (C,,  in  figure  16)  to  prevent 
bypassing  around  the  crystal  undesired  signals 
off  the  crystal  resonant  frequency.  The  bal- 
ancing is  done  by  a phasing  circuit  which 
takes  out-of-phase  voltage  from  a balanced  in- 


Inl 


Figure  18 

TYPICAL  CRYSTAL  FILTER  CIRCUIT 


put  circuit  and  passes  it  to  the  output  side  of 
the  crystal  in  proper  phase  to  neutralize  that 
passed  through  the  holder  capacitance.  A rep- 
resentative practical  filter  arrangement  is 
shown  in  figure  18.  The  balanced  input  circuit 
may  be  obtained  either  through  the  use  of  a 
split-stator  capacitor  as  shown,  or  by  the  use 
of  a center-tapped  input  coil. 

Variable-Selec-  In  the  circuit  of  figure  18,  rhe 
tivity  Filters  selectivity  is  minimum  with 
the  crystal  input  circuit  tuned 
to  resonance,  since  at  resonance  the  imped- 
ance of  the  tuned  circuit  is  maximum.  As  the 
input  circuit  is  detuned  from  resonance,  how- 
ever, the  impedance  decreases,  and  the  selec- 
tivity becomes  greater.  In  this  circuit,  the  out- 
put from  the  crystal  filter  is  tapped  down  on 
the  i-f  stage  grid  winding  to  provide  a low 
value  of  series  impedance  in  the  output  cir- 
cuit. It  will  be  recalled  that  for  maximum  selec- 
tivity, the  total  impedance  in  series  with  the 
crystal  (both  input  and  output  circuits)  must 
be  low.  If  one  is  made  low  and  the  other  is 
made  variable,  then  the  selectivity  may  be 
varied  at  will  from  sharp  to  broad. 

The  circuit  shown  in  figure  19  also  achieves 
variable  selectivity  by  adding  a variable  im- 
pedance in  series  with  the  crystal  circuit.  In 
this  case,  the  variable  impedance  is  in  series 
with  the  crystal  output  circuit.  The  impedance 
of  the  output  circuit  is  varied  by  varying  the 
Q.  As  the  Q is  reduced  (by  adding  resistance 
in  series  with  the  coil),  the  impedance  de- 
creases and  the  selectivity  becomes  greater. 
The  input  circuit  impedance  is  made  low  by 
using  a non-resonant  secondary  on  the  input 
transformer. 

A variation  of  the  circuit  shown  at  figure 
19  consists  of  placing  the  variable  resistance 
across  the  coil  and  capacitor,  rather  than  in 
series  with  them.  The  result  of  adding  the  re- 
sistor is  a reduction  of  the  output  impedance, 
and  an  increase  in  selectivity.  The  circuit  be- 
haves oppositely  to  that  of  figure  19,  however; 
as  the  resistance  is  lowered  the  selectivity 
becomes  greater.  Still  another  variation  of  fig- 
ure 19  is  to  use  the  tuning  capacitor  across 
the  output  coil  to  vary  the  output  impedance. 


222  Radio  Receiver  Fundamentals 


THE  RADIO 


CRYSTAL 


Figure  19 

VARIABLE  SELECTIVITY 
CRYSTAL  FILTER 

This  circuit  permits  of  a greater  control  of  selec- 
tivity than  does  the  circuit  of  figure  16,  and  does 
not  require  a split-stator  variable  capacitor. 


As  the  output  circuit  is  detuned  from  reso- 
nance, its  impedance  is  lowered,  and  the 
selectivity  increases.  Sometimes  a set  of 
fixed  capacitors  and  a multipoint  switch  are 
used  to  give  step-by-step  variation  of  the  out- 
put circuit  tuning,  and  thus  of  the  crystal 
filter  selectivity. 

Rejection  As  previously  discussed,  a filter 
Notch  crystal  has  both  a resonant(series 

resonant)  and  an  anti-resonant 
(parallel  resonant)  frequency,  the  impedance 
of  the  crystal  being  quite  low  at  the  former 
frequency,  and  quite  high  at  the  latter  fre- 
quency. The  anti-resonant  frequency  is  just 
slightly  higher  than  the  resonant  frequency, 
the  difference  depending  upon  the  effective 
shunt  capacitance  of  the  filter  crystal  and 
holder.  As  adjustment  of  the  phasing  capacitor 
controls  the  effective  shunt  capacitance  of  the 
crystal,  it  is  possible  to  vary  the  anti-reso- 
nant frequency  of  the  crystal  slightly  without 
unbalancing  the  circuit  sufficiently  to  let  un- 
desired signals  leak  through  the  shunt  capac- 
itance in  appreciable  amplitude.  At  the  exact 
anti-resonant  frequency  of  the  crystal  the  at- 
tenuation is  exceedingly  high,  because  of  the 
high  impedance  of  the  crystal  at  this  frequen- 
cy. This  is  called  the  rejection  notch,  and 
can  be  utilized  virtually  to  eliminate  the 
heterodyne  image  or  repeat  tuning  of  c-w  sig- 
nals. The  beat  frequency  oscillator  can  be 
so  adjusted  and  the  phasing  capacitor  so  ad- 
justed that  the  desired  beat  note  is  of  such 
a pitch  that  the  image  (the  same  audio  note 
on  the  other  side  of  zero  beat)  falls  in  the  re- 
jection notch  and  is  inaudible.  The  receiver 
then  is  said  to  be  adjusted  for  single-signal 
operation. 

The  rejection  notch  sometimes  can  be  em- 
ployed to  reduce  interference  from  an  un- 
desired phone  signal  which  is  very  close  in 
frequency  to  a desired  phone  signal.  The  filter 
is  adjusted  to  "broad”  so  as  to  permit  tele- 


0 


Figure  20 

l-F  PASS  BAND  OF  TYPICAL 
CRYSTAL  FILTER 
COMMUNICATIONS  RECEIVER 


phony  reception,  and  the  receiver  tuned  so 
that  the  carrier  frequency  of  the  undesired 
signal  falls  in  the  rejection  notch.  The  modu- 
lation sidebands  of  the  undesired  signal  still 
will  come  through,  but  the  carrier  heterodyne 
will  be  effectively  eliminated  and  interference 
greatly  reduced. 

A typical  crystal  selectivity  curve  for  a 
communications  receiver  is  shown  in  figure  20. 

Crystal  Filter  A crystal  filter,  especially 
Considerations  when  adjusted  for  single  sig- 
nal reception,  greatly  reduces 
interference  and  background  noise,  the  latter 
feature  permitting  signals  to  be  copied  that 
would  ordinarily  be  too  weak  to  be  heard  above 
the  background  hiss.  However,  when  the  filter 
is  adjusted  for  maximum  selectivity,  the  pass 
band  is  so  narrow  that  the  received  signal 
must  have  a high  order  of  stability  in  order  to 
stay  within  the  pass  band.  Likewise,  the  local 
oscillator  in  the  receiver  must  be  highly  stable, 
or  constant  retuning  will  be  requited.  Another 
effect  that  will  be  noticed  with  the  filter  ad- 
justed too  "sharp”  is  a tendency  for  code 
characters  to  produce  a ringing  sound,  and 
have  a hangover  or  "tails.”  This  effect  limits 
the  code  speed  that  can  be  copied  satisfac- 
torily when  the  filter  is  adjusted  for  extreme 
selectivity. 

The  Mechanical  The  Collins  Mechanical  FiT 
Filter  ter  (figure  21)  is  a new  con- 

cept in  the  field  of  selec- 
tivity. It  is  an  electro-mechanical  bandpass 
filter  about  half  the  size  of  a cigarette  pack- 
age. As  shown  in  figure  22,  it  consists  of  an 
input  transducer,  a resonant  mechanical  sec- 


HANDBOOK 


Collins  Mechanical  Filter  223 


tion  comprised  of  a number  of  metal  discs,  and 
an  output  transducer. 

The  frequency  characteristics  of  the  reso- 
nant mechanical  section  provide  the  almost 
rectangular  selectivity  curves  shown  in  figure 
23.  The  input  and  output  transducers  serve 
only  as  electrical  to  mechanical  coupling  de- 
vices and  do  not  affect  the  selectivity  charac- 
teristics which  are  determined  by  the  metal 
discs.  An  electrical  signal  applied  to  the  in- 
put terminals  is  converted  into  a mechanical 
vibration  at  the  input  transducer  by  means  of 
magnetostriction.  This  mechanical  vibration 
travels  through  the  resonant  mechanical  sec- 
tion to  the  output  transducer,  where  it  is  con- 
verted by  magnetostriction  to  an  electrical 
signal  which  appears  at  the  output  terminals. 

In  order  to  provide  the  most  efficient  electro- 
mechanical coupling,  a small  magnet  in  the 
mounting  above  each  transducer  applies  a mag- 
netic bias  to  the  nickel  transducer  core.  The 
electrical  impulses  then  add  to  or  subtract 
from  this  magnetic  bias,  causing  vibration  of 
the  filter  elements  that  corresponds  to  the 
exciting  signal.  There  is  no  mechanical  motion 
except  for  the  imperceptible  vibration  of  the 
metal  discs. 

Magnetostrictively-driven  mechanical  filters 
have  several  advantages  over  electrical  equi- 
valents. In  the  region  from  100  kc.  to  500  kc., 
the  mechanical  elements  rue  extremely  small, 
and  a mechanical  filter  having  better  selectivi- 
ty than  the  best  of  conventional  i-f  systems 
may  be  enclosed  in  a package  smaller  than  one 
i-f  transformer. 

Since  mechanical  elements  with  Q’s  of  5000 
or  more  are  readily  obtainable,  mechanical  fil- 
ters may  be  designed  in  accordance  with  the 
theory  for  lossless  elements.  This  permits  fil- 
ter characteristics  that  are  unobtainable  with 
electrical  circuits  because  of  the  relatively 
high  losses  in  electrical  elements  as  compared 
with  the  mechanical  elements  used  in  the 
filters. 


Figure  21 


COLLINS  MECHANICAL  FILTERS 

The  Collins  Mechanical  Filter  is  an 
electro-mechanical  bandpass  filter  which 
surpasses,  in  one  small  unit,  the  se- 
lectivity of  conventional,  space-consuming 
filters.  At  the  left  is  tne  miniaturized 
filter,  less  than  214*  long.  Type  H is 
next,  and  two  horizontal  mounting  types 
are  at  right.  For  exploded  view  of  Collins 
Mechanical  Filter,  see  figure  46. 


The  frequency  characteristics  of  the  mechan- 
ical filter  are  permanent,  and  no  adjustment  is 
required  or  is  possible.  The  filter  is  enclosed 
in  a hermetically  sealed  case. 

In  order  to  realize  full  benefit  from  the  me- 
chanical filter’s  selectivity  characteristics, 
it  is  necessary  to  provide  shielding  between 
the  external  input  and  output  circuits,  capable 
of  reducing  transfer  of  energy  external  to  the 


0 


electrical  SIGNAL  ELECTRICAL  SIGNAL 

(input  or  output)  (input  or  output) 


Figure  22 

MECHANICAL  FILTER 
FUNCTIONAL  DIAGRAM 


Figure  23 

Selectivity  curves  of  4S54c&  mecltonicai  filters 
with  nominal  0.8-kc.  (clotted  line)  and  3.1-kc. 
(solid  line)  bandwidth  at  *6  db. 


224  Radio  Receiver  Fundamentals 


THE  RADIO 


AUDIO 


Figure  24 

VARIABLE-OUTPUT  B-F-0  CIRCUIT 
A beat'frequency  oscillator  whose  output  is  con- 
trollable is  of  considerable  assistance  in  copying 
c-w  signals  over  a wide  range  of  levels^  and  such 
a control  is  almost  a rtecessity  for  satisfactory 
copying  of  single-sideband  radiophone  signals. 


filter  by  a minimum  value  of  100  db.  If  the  in- 
put circuit  is  allowed  to  couple  energy  into 
the  output  circuit  external  to  the  filter,  the 
excellent  skirt  selectivity  will  deteriorate  and 
the  passband  characteristics  will  be  distorted. 

As  with  almost  any  mechanically  resonant 
circuit,  elements  of  the  mechanical  filter  have 
multiple  resonances.  These  result  in  spurious 
modes  of  transmission  through  the  filter  and 
produce  minor  passbands  at  frequencies  on 
other  sides  of  the  primary  passband.  Design 
of  the  filter  reduces  these  sub-bands  to  a low 
level  and  removes  them  from  the  immediate 
area  of  the  major  passband.  Two  conventional 
i-f  transformers  supply  increased  attenuation 
to  these  spurious  responses,  and  are  sufficient 
to  reduce  them  to  an  insignificant  level. 

Beat- Frequency  The  beat-frequency  oscillator. 
Oscillators  usually  called  the  b.f.o,,  is  a 

necessary  adjunct  for  recep- 
tion of  c-w  telegraph  signals  on  superhetero- 
dynes which  have  no  other  provision  for  ob- 
taining modulation  of  an  incoming  c-w  tele- 
graphy signal.  The  oscillator  is  coupled  into 
or  just  ahead  of  second  detector  circuit  and 
supplies  a signal  of  nearly  the  same  frequency 
as  that  of  the  desired  signal  from  the  i-f  am- 
plifier. If  the  i-f  amplifier  is  tuned  to  455  kc., 
for  example,  the  b.f.o.  is  tuned  to  approxi- 
mately 454  or  456  kc.  to  produce  an  audible 
(1000  cycle)  beat  note  in  the  output  of  the 
second  detector  of  the  receiver.  The  carrier 
signal  itself  is,  of  course,  inaudible.  The  b.f.o. 
is  not  used  for  voice  reception,  except  as  an 
aid  in  searching  for  weak  stations. 

The  b-f-o  input  to  the  second  detector  need 
only  be  sufficient  to  give  a good  beat  note  on 
an  average  signal.  Too  much  coupling  into  the 
second  detector  will  give  an  excessively  high 
hiss  level,  masking  weak  signals  by  the  high 
noise  background. 

Figure  24  shows  a method  of  manually  ad- 


(.F,  STAGE  DET. 


@ DIODE  DETECTOR 


I.F.  STAGE  DET. 


@ PLATE  DETECTOR 


I.F.  STAGE  DET. 


INFINITE  IMPEDANCE  DETECTOR 


Figure  25 

TYPICAL  CIRCUITS  FOR  GRID-LEAK, 
DIODE,  PLATE  AND  INFINITE  IMPE- 
DANCE DETECTOR  STAGES 


justing  the  b-f-o  output  to  correspond  with  the 
strength  of  received  signals.  This  type  of  vari- 
able b-f-o  output  control  is  a useful  adjunct 
to  any  superheterodyne,  since  it  allows  suffi- 
cient b-f-o  output  to  be  obtained  to  beat  with 
strong  signals  or  to  allow  single-sideband 
reception  and  at  the  same  time  permits  the 
b-f-o  output,  and  consequently  the  hiss,  to  be 
reduced  when  attempting  to  receive  weak  sig- 
nals. The  circuit  shown  is  somewhat  better 
than  those  in  which  one  of  the  electrode  volt- 


HANDBOOK 


Detector  Circuits  225 


Figure  26 

TYPtCAL  A-V-C  CIRCUIT  USING  A DOUBLE  DIODE 

Any  of  the  small  dual-diode  tubes  may  he  used  in  this  circuit.  Or,  if  desired,  a duo-diode- 
tnode  m^  be  used,  with  the  triade  acting  as  the  first  audio  stage.  The  left-hand  diode  serves 
as  the  dete^or,  while  the  right-hand  side  acts  as  the  a-v-c  rectifier.  The  use  of  separate 
diodes  for  detector  and  a-v-c  reduces  distortion  when  receiving  an  AM  signal  with  a high 

modulation  percentage. 


ages  on  the  b-f-o  tube  is  changed,  as  the  latter 
citcujts  usually  change  the  frequency  of  the 
b.f.o.  at  the  same  time  they  change  the  strength, 
making  it  necessary  to  reset  the  trimmer  each 
time  the  output  is  adjusted. 

The  b.f.o.  usually  is  provided  with  a small 
trimmer  which  is  adjustable  from  the  front 
panel  to  permit  adjustment  over  a range  of  5 
or  10  kc.  For  single-signal  reception  the  b.f.o. 
always  is  adjusted  to  the  high-frequency  side, 
in  order  to  permit  placing  the  heterodyne  image 
in  the  rejection  notch. 

In  order  to  teduce  the  b-f-o  signal  output 
voltage  to  a reasonable  level  which  will  pre- 
vent blocking  the  second  detector,  the  signal 
voltage  is  delivered  through  a low-capacitance 
(high-reactance)  capacitor  having  a value  of 
1 to  2 fififd. 

Care  must  be  taken  with  the  b.f.o.  to  pre- 
vent harmonics  of  the  oscillator  from  being 
picked  up  at  multiples  of  the  b-f-o  frequency. 
The  complete  b.f.o.  togethet  with  the  coupling 
circuits  to  the  second  detector,  should  be  thor- 
oughly shielded  to  prevent  pickup  of  the  har- 
monics by  the  input  end  of  the  receiver. 

If  b-f-o  harmonics  still  have  a tendency  to 
give  trouble  after  complete  shielding  and  iso- 
lation of  the  b-f-o  circuit  has  been  accom- 
plished, the  passage  of  these  harmonics  from 
the  b-f-o  circuit  to  the  rest  of  the  receiver  can 
be  stopped  through  the  use  of  a low-pass  filter 
in  the  lead  between  the  output  of  the  b-f-o  cir- 
cuit and  the  point  on  the  receiver  where  the 
b-f-o  signal  is  to  be  injected. 

12-8  Detector,  Audio,  and 

Control  Circuits 

Detectors  Second  detectors  for  use  in  super- 

heterodynes are  usually  of  the 


diode,  plate,  or  infinite-impedance  types.  Oc- 
casionally, grid-leak  detectors  are  used  in  re- 
ceivers using  one  i-f  stage  or  none  at  all,  in 
which  case  the  second  detector  usually  is 
made  regenerative. 

Diodes  are  the  most  popular  second  detect- 
ors because  they  allow  a simple  method  of 
obtaining  automatic  volume  control  to  be  used. 
Diodes  load  the  tuned  circuit  to  which  they 
are  connected,  however,  and  thus  teduce  the 
selectivity  slightly.  Special  i-f  transformers 
are  used  for  the  purpose  of  providing  a low- 
impedance  input  circuit  to  the  diode  detector. 

Typical  circuits  for  grid-leak,  diode,  plate 
and  infinite-impedance  detectors  are  shown 
in  figure  25. 

Automatic  Vol-  The  elements  of  an  automatic 
ume  Control  volume  control  (a.v.c.)  sys- 
tem ate  shown  in  figure  26. 
A dual-diode  tube  is  used  as  a combination 
diode  detector  and  a-v-c  rectifier.  The  left- 
hand  diode  operates  as  a simple  tectifiet  in 
the  manner  described  earlier  in  this  chapter. 
Audio  voltage,  superimposed  on  a d-c  voltage, 
appears  across  the  500,000-ohm  potentiometer 
(the  volume  control)  and  the  .0001-pfd.  capa- 
citor, and  is  passed  on  to  the  audio  amplifier. 
The  right-hand  diode  receives  signal  voltage 
directly  from  the  primary  of  the  last  i-f  ampli- 
fier, and  acts  as  the  a-v-c  rectifier.  The  pul- 
sating d-c  voltage  across  the  1-megohm  a.v.c.- 
diode  load  resistor  is  filtered  by  a 500,000-ohm 
resistor  and  a .05-/tfd.  capacitor,  and  applied 
as  bias  to  the  grids  of  the  r-f  and  i-f  amplifier 
tubes;  an  increase  or  decrease  in  signal 
strength  will  cause  a corresponding  increase 
or  decrease  in  a-v-c  bias  voltage,  and  thus  the 
gain  of  the  receiver  is  automatically  adjusted 
to  compensate  for  changes  in  signal  strength. 


226  Radio  Receiver  Fundamentals 


THE  RADIO 


A-C  Loading  of  By  disassociating  the  a.v.c. 
Second  Detector  and  detecting  functions 
through  using  separate 
diodes,  as  shown,  most  of  the  ill  effects  of 
a-c  shunt  loading  on  the  detector  diode  are 
avoided.  This  type  of  loading  causes  serious 
distortion,  and  the  additional  components  re- 
quired to  eliminate  it  are  well  worth  their  cost. 
Even  with  the  circuit  shown,  a-c  loading  can 
occur  unless  a very  high  (5  megohms,  or  more) 
value  of  grid  resistor  is  used  in  the  following 
audio  amplifier  stage. 

A.V.C.  in  In  receivers  having  a beat- 

B-F-O- Equipped  frequency  oscillator  for  the 

Receivers  reception  of  radiotelegraph 

signals,  the  use  of  a.v.c. 
can  result  in  a great  loss  in  sensitivity  when 
the  b.f.o.  is  switched  on.  This  is  because  the 
beat  oscillator  output  acts  exactly  like  a 
strong  received  signal,  and  causes  the  a-v-c 
circuit  to  put  high  bias  on  the  r-fandi-f  stages, 
thus  greatly  reducing  the  receiver ’s  sensitivi- 
ty. Due  to  the  above  effect,  it  is  necessary  to 
provide  a method  of  making  the  a-v-c  circuit 
inoperative  when  the  b.f.o.  is  being  used.  The 
simplest  method  of  eliminating  the  a-v-c  ac- 
tion is  to  short  the  a-v-c  line  to  ground  when 
the  b.f.o.  is  turned  on.  A two-circuit  switch 
may  be  used  for  the  dual  purpose  of  turning  on 
the  beat  oscillator  and  shorting  out  the  a.v.c. 
if  desired. 

Signal  Strength  Visual  means  for  determining 
Indicators  whether  or  not  the  receiver  is 

properly  tuned,  as  well  as  an 
indication  of  the  relative  signal  strength,  ate 
both  provided  by  means  of  tuning  indicators 
(S  meters)  of  the  meter  or  vacuum-tube  type. 
A d-c  milliammeter  can  be  connected  in  the 
plate  supply  circuit  of  one  or  more  r-f  or  i-f 
amplifiers,  as  shown  in  figure  27A,  so  that  the 
change  in  plate  current,  due  to  the  action  of 
the  a-v-c  voltage,  will  be  indicated  on  the  in- 
strument. The  d-c  instrument  MA  should  have 
a full-scale  reading  approximately  equal  to  the 
total  plate  current  taken  by  the  stage  or  stages 
whose  plate  current  passes  through  the  instru- 
ment. The  value  of  this  current  can  be  esti- 
mated by  assuming  a plate  current  on  each 
stage  (with  no  signal  input  to  the  receiver)  of 
about  6 ma.  However,  it  will  be  found  to  be 
more  satisfactory  to  measure  the  actual  plate 
current  on  the  stages  with  a milliammeter  of 
perhaps  0-100  ma.  full  scale  before  purchasing 
an  instrument  for  use  as  an  S meter.  The 
50-ohm  potentiometer  shown  in  the  drawing  is 
used  to  adjust  the  meter  reading  to  full  scale 
with  no  signal  input  to  the  receiver. 

When  an  ordinary  meter  is  used  in  the  plate 
circuit  of  a stage,  for  the  purpose  of  indicat- 
ing signal  strength,  the  meter  reads  backwards 


RF  IF  IF 


Figure  27 

SIGNAL-STRENGTH-METER  CIRCUITS 

Shorni  above  ore  four  circuits  for  obtaining  a sig- 
nal’Strengtb  reading  which  is  a function  of  incom- 
ing carrier  amplitude-  The  circuits  are  discussed 
tin  the  accompanying  text- 


with  respect  to  strength.  This  is  because  in- 
creased a-v-c  bias  on  stronger  signals  causes 
lower  plate  current  through  the  meter.  For  this 
reason,  special  meters  which  indicate  zero  at 
the  right-hand  end  of  the  scale  are  often  used 
for  signal  strength  indicators  in  commercial 
receivers  using  this  type  of  circuit.  Alter- 
natively, the  meter  may  be  mounted  upside 
down,  so  that  the  needle  moves  toward  the 
tight  with  increased  strength. 

The  circuit  of  figure  27B  can  frequently  be 
used  to  advantage  in  a receiver  where  the  cath- 
ode of  one  of  the  r-f  or  i-f  amplifier  stages 
runs  directly  to  ground  through  the  cathode 
bias  resistor  instead  of  running  through  a cath- 


HANDBOOK 


Noise  Suppression  227 


ode-voltage  gain  control.  In  this  case  a 0-1 
d-c  milliammeter  in  conjunction  with  a resistor 
from  1000  to  3000  ohms  can  be  used  as  shown 
as  a signal- strength  meter.  With  this  circuit 
the  meter  will  read  backwards  with  increasing 
signal  strength  as  in  the  circuit  previously 
discussed. 

Figure  27C  is  the  circuit  of  a forward-read- 
ing S meter  as  is  often  used  in  communications 
receivers.  The  instrument  is  used  in  an  un- 
balanced bridge  circuit  with  the  d-c  plate  re- 
sistance of  one  i-f  tube  as  one  leg  of  the 
bridge  and  with  resistors  for  the  other  three 
legs.  The  value  of  the  resistor  R must  be  de- 
termined by  trial  and  error  and  will  be  some- 
where in  the  vicinity  of  50,000  ohms.  Some- 
times the  screen  circuits  of  the  t-f  and  i-f 
stages  ate  taken  from  this  point  along  with 
the  screen-circuit  voltage  divider. 

Electron-ray  tubes  (sometimes  called  '*magic 
eyes”)  can  also  be  used  as  indicators  of  rela- 
tive signal  strength  in  a circuit  similar  to  that 
shown  in  figure  27D.  A 6U5/6G5  tube  should 
be  used  where  the  a-v-c  voltage  will  be  from 
5 to  20  volts  and  a type  6E5  tube  should  be 
used  when  the  a-v-c  voltage  will  run  from  2 
to  8 volts. 

Audio  Amplifiers  Audio  amplifiers  are  em- 
ployed in  neatly  all  radio 
receivers.  The  audio  amplifier  stage  or  stages 
ate  usually  of  the  Class  A type,  although  Class 
AB  push-pull  stages  are  used  in  some  re- 
ceivers. The  purpose  of  the  audio  amplifier  is 
to  bring  the  relatively  weak  signal  from  the 
detector  up  to  a strength  sufficient  to  operate 
a pair  of  headphones  or  a loud  speaker.  Either 
triodes,  pentodes,  or  beam  tetrodes  may  be 
used,  the  pentodes  and  beam  tetrodes  usually 
giving  greater  output.  In  some  receivers,  par- 
ticularly those  employing  grid  leak  detection, 
it  is  possible  to  operate  the  headphones  di- 
rectly from  the  detector,  without  audio  ampli- 
fication. In  such  receivers,  a single  audio 
stage  with  a beam  tetrode  or  pentode  tube  is 
ordinarily  used  to  drive  the  loud  speaker. 

Most  communications  receivers,  either  home- 
constructed  or  factory-made,  have  a single- 
ended  beam  tetrode  (such  as  a 6L6  or  6V6)  or 
pentode  (6F6  or  6K6-GT)  in  the  audio  output 
stage  feeding  the  loudspeaker.  If  precautions 
are  not  taken  such  a stage  will  actually  bring 
about  a decrease  in  the  effective  signal-to- 
noise  ratio  of  the  receiver  due  to  the  rising 
high-frequency  characteristic  of  such  a stage 
when  feeding  a loud-speaker.  One  way  of  im- 
proving this  condition  is  to  place  a mica  or 
paper  capacitor  of  approximately  0.003  pfd. 
capacitance  across  the  primary  of  the  output 
transformer.  The  use  of  a capacitor  in  this 
manner  tends  to  make  the  load  impedance  seen 
by  the  plate  of  the  output  tube  more  constant 


over  the  audio-frequency  range.  The  speaker 
and  transformer  will  tend  to  present  a rising 
impedance  to  the  tube  as  the  frequency  in- 
creases, and  the  parallel  capacitor  will  tend 
to  make  the  total  impedance  more  constant 
since  it  will  tend  to  present  a decreasing  im- 
pedance with  increasing  audio  frequency. 

A still  better  way  of  improving  the  frequency 
characteristic  of  the  output  stage,  and  at  the 
same  time  reducing  the  harmonic  distortion, 
is  to  use  shunt  feedback  from  the  plate  of  the 
output  tube  to  the  plate  of  a tube  such  as  a 
6SJ7  acting  as  an  audio  amplifier  stage  ahead 
of  the  output  stage. 


12-9  Noise  Suppression 

The  problem  of  noise  suppression  confronts 
the  listener  who  is  located  in  places  where 
interference  from  power  lines,  electrical  ap- 
pliances, and  automobile  ignition  systems  is 
troublesome.  This  noise  is  often  of  such  in- 
tensity as  to  swamp  out  signals  from  desired 
stations. 

There  are  two  principal  methods  for  reduc- 
ing this  noise; 

(1)  A-c  line  filters  at  the  source  of  inter- 
ference, if  the  noise  is  created  by  an 
electrical  appliance. 

(2)  Noise-limiting  circuits  for  the  reduc- 
tion, in  the  receiver  itself,  of  interfer- 
ence of  the  type  caused  by  automobile 
ignition  systems. 

Power  Line  Many  household  appliances,  such 
Filters  as  electric  mixers,  heating  pads, 

vacuum  sweepers,  refrigerators, 
oil  burners,  sewing  machines,  doorbells,  etc., 
create  an  interference  of  an  intermittent  na- 
ture. The  insertion  of  a line  filter  near  the 
source  of  interference  often  will  effect  a com- 
plete cure.  Filters  for  small  appliances  can 
consist  of  a 0.1-/xfd.  capacitor  connected  a- 
cross  the  110-volt  a-c  line.  Two  capacitors 
in  series  across  the  line,  with  the  midpoint 
connected  to  ground,  can  be  used  in  conjunc- 
tion with  ultraviolet  ray  machines,  refriger- 
ators, oil  burner  furnaces,  and  other  more  stub- 
born offenders.  In  severe  cases  of  interfer- 
ence, additional  filters  in  the  form  of  heavy- 
duty  r-f  choke  coils  must  be  connected  in 
series  with  the  110- volt  a-c  line  on  both  sides 
of  the  line  right  at  the  interfering  appliance. 

Peak  Noise  Numerous  noise-limiting  circuits 

Limiters  which  are  beneficial  in  overcom- 

ing key  clicks,  automobile  ig- 
nition interference,  and  similar  noise  impulses 
have  become  popular.  They  operate  on  the 
principle  that  each  individual  noise  pulse  is 


228  Radio  Receiver  Fundamentals 


THE  RADIO 


of  very  short  duration,  yet  of  very  high  ampli- 
tude. The  popping  or  clicking  type  of  noise 
from  electrical  ignition  systems  may  produce 
a signal  having  a peak  value  ten  to  twenty 
times  as  great  as  the  incoming  radio  signal, 
but  an  average  power  much  less  than  the  sig- 
nal. 

As  the  duration  of  this  type  of  noise  peak 
is  short,  the  receiver  can  be  made  inoperative 
during  the  noise  pidse  without  the  human  ear 
detecting  the  total  loss  of  signal.  Some  noise 
limiters  actually  punch  a hole  in  the  signal, 
while  others  merely  limit  the  maximum  peak 
signal  which  reaches  the  headphones  or  loud- 
speaker. 

The  noise  peak  is  of  such  short  duration 
that  it  would  not  be  objectionable  except  for 
the  fact  that  it  produces  an  over-loading  effect 
on  the  receiver,  which  increases  its  time  con- 
stant. A sharp  voltage  peak  will  give  a kick 
to  the  diaphragm  of  the  headphones  or  speak- 
er, and  the  momentum  or  inertia  keeps  the 
diaphragm  in  motion  until  the  dampening  of 
the  diaphragm  stops  it.  This  movement  pro- 
duces a popping  sound  which  may  completely 
obliterate  the  desired  signal.  If  the  noise  pulse 
can  be  limited  to  a peak  amplitude  equal  to 
that  of  the  desired  signal,  the  resulting  inter- 
ference is  practically  negligible  for  moder- 
ately low  repetition  rates,  such  as  ignition 
noise. 

In  addition,  the  i-f  amplifier  of  the  receiver 
will  also  tend  to  lengthen  the  duration  of  the 
noise  pulses  because  the  relatively  high-Q  i-f 
tuned  circuits  will  ring  or  oscillate  when  ex- 
cited by  a sharp  pulse,  such  as  produced  by 
ignition  noise.  The  most  effective  noise  limiter 
would  be  placed  before  the  high-Q  i-f  tuned 
circuits.  At  this  point  the  noise  pulse  is  the 
sharpest  and  has  not  been  degraded  by  pass- 
age through  the  i-f  transformers.  In  addition, 
the  pulse  is  eliminated  before  it  can  produce 
ringing  effects  in  the  i-f  chain. 


The  Lamb  An  i-f  noise  limiter  is  shown  in 

Noise  Limiter  figure  28.  This  is  an  adapta- 
tion of  the  Lamb  noise  silenc- 
er circuit.  The  i-f  signal  is  fed  into  a double 
grid  tube,  such  as  a 6L7,  and  thence  into  the 
i-f  chain.  A 6AB7  high  gain  pentode  is  capa- 
city coupled  to  the  input  of  the  i-f  system. 
This  auxiliary  tube  amplifies  both  signal  and 
noise  that  is  fed  to  it.  It  has  a minimum  of 
selectivity  ahead  of  it  so  that  it  receives  the 
true  noise  pulse  before  it  is  degraded  by  the 
i-f  strip.  A broadly  tuned  i-f  transformer  is  used 
to  couple  the  noise  amplifier  to  a 6H6  noise 
rectifier.  The  gain  of  the  noise  amplifier  is 
controlled  by  a potentiometer  in  the  cathode 
of  the  6AB7  noise  amplifier.  This  potenti- 
ometer controls  the  gain  of  the  noise  amplifier 


ISTDET.  l5TI,f,  ENOl.F, 


6L7 


Figure  28 

THE  LAMB  I-F  NOISE  SILENCER 


Stage  and  in  addition  sets  the  bias  level  on 
the  6H6  diode  so  that  the  incoming  signal  will 
not  be  rectified.  Only  noise  peaks  louder  than 
the  signal  can  overcome  the  testing  bias  of 
the  6H6  and  cause  it  to  conduct.  A noise  pulse 
rectified  by  the  6H6  is  applied  as  a negative 
voltage  to  the  control  grid  of  the  6L7  i-f  tube, 
disabling  the  tube,  and  punching  a hole  in  the 
signal  at  the  instant  of  the  noise  pulse.  By 
varying  the  bias  control  of  the  noise  limiter, 
the  negative  control  voltage  applied  to  the 
6L7  may  be  adjusted  until  it  is  barely  suffi- 
cient to  overcome  the  noise  impulses  applied 
to  the  #1  control  grid  without  allowing  the 
modulation  peaks  of  the  carrier  to  become 
badly  distorted. 

The  Bishop  Another  effective  i-f  noise 
Noise  Limiter  limiter  is  the  Bishop  limiter. 

This  is  a full-wave  shunt  type 
diode  limiter  applied  to  the  primary  of  the  last 
i-f  transformer  of  a receiver.  The  limiter  is 
self-biased  and  automatically  adjusts  itself 
to  the  degree  of  modulation  of  the  received 
signal.  The  schematic  of  this  limiter  is  shown 
in  figure  29.  The  bias  circuit  time  constant  is 
determined  by  Ci  and  the  shunt  resistance, 
which  consists  of  Rj  and  Rj  in  series.  The 
plate  resistance  of  the  last  i-f  tube  and  the 
capacity  of  C,  determine  the  charging  rate  of 
the  circuit.  The  limiter  is  disabled  by  opening 
S„  which  allows  the  bias  to  rise  to  the  value 
of  the  i-f  signal. 


HANDBOOK 


Noise  Limiters  229 


Audio  Noise  Some  of  the  simplest  and  most 
Limiters  practical  peak  limiters  for  radio- 

telephone reception  employ  one 
or  two  diodes  either  as  shunt  or  series  limiters 
in  the  audio  system  of  the  receiver.  When  a 
noise  pulse  exceeds  a certain  predetermined 
threshold  value,  the  limiter  diode  acts  either 
as  a short  or  open  circuit,  depending  upon 
whether  it  is  used  in  a shunt  or  series  circuit. 
The  threshold  is  made  to  occur  at  a level  high 
enough  that  it  will  not  clip  modulation  peaks 
enough  to  impair  voice  intelligibility,  but  low 
enough  to  limit  the  noise  peaks  effectively. 

Because  the  action  of  the  peak  limiter  is 
needed  most  on  very  weak  signals,  and  these 
usually  ate  not  strong  enough  to  produce  proper 
a-v-c  action,  a threshold  setting  that  is  cor- 
rect for  a strong  phone  signal  is  not  correct 
for  optimum  limiting  on  very  weak  signals.  For 
this  reason  the  threshold  control  often  is  tied 
in  with  the  a-v-c  system  so  as  to  make  the 
optimum  threshold  adjustment  automatic  in- 
stead of  manual. 

Suppression  of  impulse  noise  by  metuis  of 
an  audio  peak  limiter  is  best  accomplished  at 
the  very  front  end  of  the  audio  system,  and 
for  this  reason  the  function  of  superheterodyne 
second  detector  and  limiter  often  are  combined 
in  a composite  circuit. 

The  amount  of  limiting  that  can  be  obtained 
is  a function  of  the  audio  distortion  that  can 
be  tolerated.  Because  excessive  distortion 
will  reduce  the  intelligibility  as  much  as  will 
background  noise,  the  degree  of  limiting  for 
which  the  circuit  is  designed  has  to  be  a com- 
promise. 

Peak  noise  limiters  working  at  the  second 
detector  are  much  more  effective  when  the  i-f 
bandwidth  of  the  receiver  is  broad,  because  a 
sharp  i-f  amplifier  will  lengthen  the  pulses  by 
the  time  they  reach  the  second  detector,  mak- 
ing the  limiter  less  effective.  V-h-f  super- 
heterodynes have  an  i-f  bandwidth  consider- 
ably wider  than  the  minimum  necessary  for 

voice  sidebands  (to  take  cate  of  drift  and  in- 
stability). Therefore,  they  are  capable  of  bet- 


ter peak  noise  suppression  than  a standard 
communications  receiver  having  an  i-f  band- 
width of  perhaps  8 kc.  Likewise,  when  a crys- 
tal filter  is  used  on  the  "sharp”  position  an 
a-f  peak  limiter  is  of  little  benefit. 

Practical  Noise  limiters  range  all  the 

Peak  Noise  way  from  an  audio  stage 

Limiter  Circuits  running  at  very  low  screen 
or  plate  voltage,  to  elabor- 
ate affairs  employing  5 or  more  tubes.  Rather 
than  attempt  to  sbow  the  numerous  types,  many 
of  which  ate  quite  complex  considering  the 
results  obtained,  only  two  very  similar  types 
will  be  described.  Either  is  just  about  as  ef- 
fective as  the  most  elaborate  limiter  that  can 
be  constructed,  yet  requires  the  addition  of 
but  a single  diode  and  a few  resistors  and 
capacitors  over  what  would  be  employed  in  a 
good  superheterodyne  without  a limiter.  Both 
circuits,  with  but  minor  modifications  in  re- 
sistance and  capacitance  values,  are  incor- 
porated in  one  form  or  another  in  different 
types  of  factory-built  communications  receiv- 
ers. 

Referring  to  figure  30,  the  first  circuit  shows 
a conventional  superheterodyne  second  de- 
tector, a.v.c.,  and  first  audio  stage  with  the 
addition  of  one  tube  element,  Dj,  which  may 
be  either  a separate  diode  or  part  of  a twin- 
diode  as  illustrated.  Diode  D,  acts  as  a series 
gate,  allowing  audio  to  get  to  the  grid  of  the 
a-f  tube  only  so  long  as  the  diode  is  conduct- 
ing. The  diode  is  biased  by  a d-c  voltage  ob- 
tained in  the  same  manner  as  a-v-c  control  volt- 
age, the  bias  being  such  that  pulses  of  short 
duration  no  longer  conduct  when  the  pulse 
voltage  exceeds  the  carrier  by  approximately 
60  pet  cent.  This  also  clips  voice  modulation 
peaks,  but  not  enough  to  impair  intelligibility. 

It  is  apparent  that  the  series  diode  clips 
only  positive  modulation  peaks,  by  limiting 
upward  modulation  to  about  60  pet  cent.  Neg- 
ative or  downward  peaks  are  limited  automati- 
cally to  100  pet  cent  in  the  detector,  because 
obviously  the  rectified  voltage  out  of  the 
diode  detector  cannot  be  less  than  zero.  Limit- 
ing the  downward  peaks  to  60  per  cent  or  so 
instead  of  100  per  cent  would  result  in  but 
little  improvement  in  noise  reduction,  and  the 
results  do  not  justify  the  additional  compon- 
ents required. 

It  is  important  that  the  exact  resistance 
values  shown  be  used,  for  best  results,  and 
that  10  per  cent  tolerance  resistors  be  used 
for  Rj  and  R,.  Also,  the  rectified  carrier  volt- 
age developed  across  Cj  should  be  at  least  5 
volts  for  good  limiting. 

The  limiter  will  work  well  on  c-w  telegraphy 
if  the  amplitude  of  beat  frequency  oscillator 
injection  is  not  too  high.  Variable  injection  is 
to  be  preferred,  adjustable  from  the  front  panel. 


230  Radio  Receiver  Fundamentals 


THE  RADIO 


V2 


Q— ‘O.U/lfd.  paper 
C^^50-/x/ifd.  mica 
C3— 100-/i'/i-fd.  mica 
04,  C5 — O.OI-Aifd.  paper 
Rl»  R2““^  megohm, 

H watt 

R3,  R4~220,000  ohms, 
watt 

R5  , Rfi— “1  megohm, 

H watt 

R7*^2-megohm  poten- 
tiometer 


Figure  30 

NOISE  LIMITER  CIRCUIT,  WITH  ASSOCIATED  A-V-C 

This  limiter  is  of  the  series  type,  and  is  self^adjustir^g  to  carrier  strength  for  phone  reception.  For  proper 
operatior}  several  volts  should  be  developed  across  the  secondary  of  the  last  i-f  transformer  (IFT)  under 

carrier  conditions. 


If  this  feature  is  not  provided,  the  b-f-o  injec- 
tion should  be  reduced  to  the  lowest  value  that 
will  give  a satisfactory  beat.  When  this  is 
done,  effective  limiting  and  a good  beat  can 
be  obtained  by  proper  adjustment  of  the  r-f 
and  a-f  gain  controls.  It  is  assumed,  of  course, 
that  the  a.v.c.  is  cut  out  of  the  circuit  for 
c-w  telegraphy  reception. 

Alternative  The  circuit  of  figure  31  is 

Limiter  Circuit  more  effective  than  that 
shown  in  figure  30  under  cer- 
tain conditions  and  requites  the  addition  of 
only  one  more  resistor  and  one  mote  capaci- 
tor than  the  other  circuit.  Also,  this  circuit 
involves  a smaller  loss  in  output  level  than 
the  circuit  of  figure  30.  This  circuit  can  be 
used  with  equal  effectiveness  with  a com- 
bined diode-triode  or  diode-pentode  tube  (6R7, 
6SR7,  6Q7,  6SQ7  or  similar  diode- triodes,  or 
6B8,  6SF7,  or  similar  diode-pentodes)  as  diode 
detector  and  first  audio  stage.  However,  a 
separate  diode  must  be  used  for  the  noise 
limiter,  Dj.  This  diode  may  be  one-half  of  a 
6H6,  6AL5,  7A6,  etc.,  or  it  may  be  a triode 
connected  6J5,  6C4  or  similar  type. 

Note  that  the  return  for  the  volume  control 
must  be  made  to  the  cathode  of  the  detector 
diode  (and  not  to  ground)  when  a dual  tube  is 
used  as  combined  second-detector  first-audio. 
This  means  that  in  the  circuit  shown  in  figure 
31  a connection  will  exist  across  the  points 
where  the  "X”  is  shown  on  the  diagram  since 
a common  cathode  lead  is  brought  out  of  the 
tube  for  D,  and  V,.  If  desired,  of  course,  a 
single  dual  diode  may  be  used  for  D,  and  Dj 
in  this  circuit  as  well  as  in  the  circuit  of  fig- 
ure 30.  Switching  the  limiter  in  and  out  with 
the  switch  S brings  about  no  change  in  volume. 

In  any  diode  limiter  circuit  such  as  the  ones 
shown  in  these  two  figures  it  is  important  that 


the  mid-point  of  the  heater  potential  for  the 
noise-limiter  diode  be  as  close  to  ground 
potential  as  possible.  This  means  that  the 
center-tap  of  the  heater  supply  for  the  tubes 
should  be  grounded  wherever  possible  rather 
than  grounding  one  side  of  the  heater  supply 
as  is  often  done.  Difficulty  with  hum  pickup 
in  the  limiter  circuit  may  be  encountered  when 
one  side  of  the  heater  is  grounded  due  to  the 
high  values  of  resistance  necessary  in  the 
limiter  circuit. 

The  circuit  of  figure  31  has  been  used  with 
excellent  success  in  several  home-constructed 
receivers,  and  in  the  BC-312/BC-342  and  BC- 
348  series  of  surplus  communications  receiv- 
ers. It  is  also  used  in  certain  manufactured 
receivers. 

An  excellent  check  on  the  operation  of  the 
noise  limiter  in  any  communications  receiver 
can  be  obtained  by  listening  to  the  Loran  sig- 
nals in  the  160-meter  band.  With  the  limiter 
out  a sharp  rasping  buzz  will  be  obtained 
when  one  of  these  stations  is  tuned  in.  With 
the  noise  limiter  switched  into  the  circuit  the 
buzz  should  be  greatly  reduced  and  a low- 
pitched  hum  should  be  heard. 

The  Full-Wave  The  most  satisafctory  diode 
Limiter  noise  limiter  is  the  series  full- 

wave  limiter,  shown  in  figure 
32.  The  positive  noise  peaks  are  clipped  by 
diode  A,  the  clipping  level  of  which  may  be 
adjusted  to  clip  at  any  modulation  level  be- 
tween 25  per  cent  and  100  per  cent.  The  nega- 
ative  noise  peaks  are  clipped  by  diode  B at 
a fixed  level. 

The  TNS  Limiter  The  Twin  Noise  Squelch, 
popularized  by  CQ  maga- 
zine, is  a combination  of  a diode  noise  clipper 
and  an  audio  squelch  tube.  Tbe  squelch  cir- 


HANDBOOK 


U-H-F  Circuits  231 


This  circuit  is  of  the  self- 
adjusting  type  and  gives  less 
distortion  for  a given  degree 
of  modulation  than  the  more 
common  limiter  circuits. 

Rif  R2~—470K,  Vi  watt 
R3  — 100k,  Vi  watt 
R4,  R5  —1  megohm,  Vi  watt 
Rg— 2-megohm  potentiometer 
C3 ->-0.00025  mica  (approx.) 

0.01-/J^fd.  paper 
C3.-.0.01-)Ufd.  paper 
C4— 0.01-/xfd.  paper 

D2— -6H6,  6AL5,  7A6,  or 

diode  sections  of  o 6S8-GT 


IFT  Dl 


Figure  31 

ALTERNATIVE  NOISE  LIMITER  CIRCUIT 


cuit  is  useful  in  eliminating  the  grinding  back- 
ground noise  that  is  the  residual  left  by  the 
diode  clipper.  In  figure  33,  the  setting  of  the 
470k  potentiometer  determines  the  operating 
level  of  the  squelch  action  and  should  be  set 
to  eliminate  the  residual  background  noise. 
Because  of  the  low  inherent  distortion  of  the 
TNS,  it  may  be  left  in  the  circuit  at  all  times. 
As  with  other  limiters,  the  TNS  requites  a 
high  signal  level  at  the  second  detector  for 
maximum  limiting  effect. 

12-10  Special  Considerations 
in  U-H-F  Receiver  Design 

Transmission  At  increasingly  higher  frequen- 

Line  Circuits  cies,  it  becomes  progressively 

more  difficult  to  obtain  a satis- 
factory amount  of  selectivity  and  impedance 
from  an  ordinary  coil  and  capacitor  used  as  a 
resonant  circuit.  On  the  other  hand,  quarter 


Figure  32 

THE  FULL-WAVE  SERIES  AUDIO 
NOISE  LIMITER 


wavelength  sections  of  parallel  conductors  or 
concentric  transmission  line  are  not  only  more 
efficient  but  also  become  of  practical  dimen- 
sions. 


Tuning  Tubes  and  tuning  capacitors  con- 

Short  Lines  nected  to  the  open  end  of  a trans- 
mission line  provide  a capaci- 
tance that  makes  the  resonant  length  less  than 
a quarter  wave-length.  The  amount  of  shorten- 
ing for  a specified  capacitive  reactance  is 
determined  by  the  surge  impedance  of  the  line 


Figure  33 

THE  TNS  AUDIO  NOISE  LIMITER 


232  Radio  Receiver  Fundamentals 


THE  RADIO 


@ ® 


Figure  34 

COUPLING  AN  ANTENNA  TO  A 
COAXIAL  RESONANT  CIRCUIT 
(A)  shows  the  reconvnended  method  for  coupling 
a coaxial  line  to  a coaxial  resonant  circuit.  181 
shows  an  alternative  method  for  use  with  an  open- 
wire  type  of  antenna  feed  line. 


Figure  35 

METHODS  OF  EXCITING  A RESONANT 
CAVITY 


section.  It  is  given  by  the  equation  for  reso- 
nance; 

1 


in  which  n = 3.1416,  / is  the  frequency,  C the 
capacitance,  Zg  the  surge  impedance  of  the 
line,  and  tan  I is  the  tangent  of  the  electrical 
length  in  degrees. 

The  capacitive  reactance  of  the  capacitance 
across  the  end  is  l/(27r  / C)  ohms.  For  reson- 
ance, this  must  equal  the  surge  impedance  of 
the  line  times  the  tangent  of  its  electrical 
length  (in  degrees,  where  90“  equals  a quar- 
ter wave).  It  will  be  seen  that  twice  the  capac- 
itance will  resonate  a line  if  its  surge  imped- 
ance is  halved;  also  that  a given  capacitance 
has  twice  the  loading  effect  when  the  frequen- 
cy is  doubled. 

Coupling  Into  It  is  possible  to  couple  into 

Lines  and  a parallel-rod  line  by  tapp- 

Cooxial  Circuits  ing  directly  on  one  or  both 
tods,  preferably  through 
blocking  capacitors  if  any  d.c.  is  present. 
Mote  commonly,  however,  a hairpin  is  induc- 
tively coupled  at  tbe  shotting  bar  end,  either 
to  the  bat  or  to  the  two  rods,  or  both.  This 
normally  will  result  in  a balanced  load.  Should 
a loop  unbalanced  to  ground  be  coupled  in, 
any  resulting  unbalance  reflected  into  the  rods 
can  be  reduced  with  a simple  Faraday  screen, 
made  of  a few  parallel  wires  placed  between 
the  hairpin  loop  and  the  rods.  These  should 
be  soldered  at  only  one  end  and  grounded. 

An  unbalanced  tap  on  a coaxial  resonant 


circuit  can  be  made  directly  on  the  inner  con- 
ductor at  the  point  where  it  is  properly  matched 
(figure  34).  For  low  impedances,  such  as  a con- 
centric line  feeder,  a small  one-half  turn  loop 
can  be  inserted  through  a hole  in  the  outer 
conductor  of  the  coaxial  circuit,  being  in  ef- 
fect a half  of  the  hairpin  type  recommended 
for  coupling  balanced  feeders  to  coaxial  re- 
sonant lines.  The  size  of  the  loop  and  close- 
ness to  the  inner  conductor  determines  the 
impedance  matching  and  loading.  Such  loops 
coupled  in  near  the  shorting  disc  do  not  alter 
the  tuning  appreciably,  if  not  ovetcoupled. 

Resonant  A cavity  is  a closed  resonant 
Cavities  chamber  made  of  metal.  It  is 
known  also  as  a rhumhatron.  The 
cavity,  having  both  inductance  and  capaci- 
tance, supersedes  coil-capacitor  and  capaci- 
tance-loaded transmission-line  tuned  circuits 
at  extremely  high  frequencies  where  conven- 
tional L and  C components,  of  even  the  most 
refined  design,  prove  impractical  because  of 
the  tiny  electrical  and  physical  dimensions 
they  must  have.  Microwave  cavities  have  high 
Q factors  and  are  superior  to  conventional 
tuned  circuits.  They  may  be  employed  in  tbe 
manner  of  an  absorption  wavemeter  or  as  the 
tuned  circuit  in  other  t-f  test  instruments,  and 
in  microwave  transmitters  and  receivers. 

Resonant  cavities  usually  are  closed  on  all 
sides  and  all  of  their  walls  are  made  of  elec- 
trical conductor.  However,  in  some  forms, 
small  openings  are  present  for  the  purpose  of 
excitation.  Cavities  have  been  produced  in 
several  shapes  including  the  plain  sphere, 


HANDBOOK 


U-H-F  Circuits  23  3 


Figure  36 

TUNING  METHODS  FOR  CYLINDRICAL 
RESONANT  CAVITIES 


dimpled  sphere,  sphere  with  reentrant  cones 
of  various  sorts,  cylinder,  prism  (including 
cube),  ellipsoid,  ellipsoid-hyperboloid,  dough- 
nut- shape,  and  various  reentrant  types.  In  ap- 
pearance, they  resemble  in  their  simpler  forms 
metal  boxes  or  cans. 

The  cavity  actually  is  a linear  circuit,  but 
one  which  is  superior  to  a conventional  co- 
axial resonator  in  the  s-h-f  range.  The  cavity 
resonates  in  much  the  same  manner  as  does  a 
barrel  or  a closed  room  with  reflecting  walls. 

Because  electromagnetic  energy,  and  the 
associated  electrostatic  energy,  oscillates  to 
and  fro  inside  them  in  one  mode  or  another, 
resonant  cavities  resemble  wave  guides.  The 
mode  of  operation  in  a cavity  is  affected  by 
the  manner  in  which  micro-wave  energy  is  in- 
jected. A cavity  will  resonate  to  a large  num- 
ber of  frequencies,  each  being  associated  with 
a particular  mode  or  standing-wave  pattern. 
The  lowest  mode  (lowest  frequency  of  opera- 
tion) of  a cavity  resonator  normally  is  the  one 
used. 

The  resonant  frequency  of  a cavity  may  be 
varied,  if  desired,  by  means  of  movable  plung- 
ers or  plugs,  as  shown  in  figure  36A,  or  a 
movable  metal  disc  (see  figure  36B).  A cavity 
that  is  too  small  for  a given  wavelength  will 
not  oscillate. 

The  resonant  frequencies  of  simple  spheri- 
cal, cylindrical,  and  cubical  cavities  may  be 
calculated  simply  for  one  particular  mode. 
Wavelength  and  cavity  dimensions  (in  centi- 
meters) are  related  by  the  following  simple 
resonance  formulae: 

For  Cylinder  At  = 2.6  x radius 
” Cube  Aj  = 2.83  X half  of  1 side 

” Sphere  A,  = 2.28x  radius 

Butterfly  Unlike  the  cavity  resonator,  which 
Circuit  in  its  conventional  form  is  a device 

which  can  tune  over  a relatively 

narrow  band,  the  butterfly  circuit  is  a tunable 
resonator  which  permits  coverage  of  a fairly 


Figure  37 

THE  BUTTERFLY  RESONANT  CIRCUIT 
Shown  at  (A)  is  the  physical  appearance  of  the 
butterfly  circuit  as  used  in  the  v-h-f  and  lower 
u-h-f  range.  (B)  shows  an  electrical  representa- 
tion of  the  circuit. 


wide  u-h-f  band.  The  butterfly  circuit  is  very 
similar  to  a conventional  coil-variable  capaci- 
tor combination,  except  that  both  inductance 
and  capacitance  ate  provided  by  what  appears 
to  be  a variable  capacitor  alone.  The  Q of  this 
device  is  somewhat  less  than  that  of  a con- 
centric-line tuned  circuit  but  is  entirely  ade- 
quate for  numerous  applications. 

Figure  37 A shows  construction  of  a single 
butterfly  section.  The  butterfly- shaped  rotor, 
from  which  the  device  derives  its  name,  turns 
in  relation  to  the  unconventional  stator.  The 
two  groups  of  stator  "fins”  or  sectors  are  in 
effect  joined  together  by  a semi-circular  metal 
band,  integral  with  the  sectors,  which  provides 
the  circuit  inductance.  When  the  rotor  is  set 
to  fill  the  loop  opening  (the  position  in  which 
it  is  shown  in  figure  37 A),  the  circuit  induct- 
ance and  capacitance  are  reduced  to  minimum. 
When  the  rotor  occupies  the  position  indicated 
by  the  dotted  lines,  the  inductance  and  capaci- 
tance are  at  maximum.  The  tuning  range  of 
practical  butterfly  circuits  is  in  the  ratio  of 

1.5:1  to  3.5:1. 

Direct  circuit  connections  may  be  made  to 
points  A and  B.  If  balanced  operation  is  de- 
sired, either  point  C or  D will  provide  the 
electrical  mid-point.  Coupling  may  be  effected 
by  means  of  a small  singleturn  loop  placed 
near  point  E or  F.  The  butterfly  thus  permits 
continuous  variation  of  both  capacitance  and 
inductance,  as  indicated  by  the  equivalent 
circuit  in  figure  37B,  while  at  the  same  time 
eliminating  all  pigtails  and  wiping  contacts. 

Several  butterfly  sections  may  be  stacked 
in  parallel  in  the  same  way  that  variable  ca- 
pacitors are  built  up.  In  stacking  these  sec- 
tions, the  effect  of  adding  inductances  in  paral- 
lel is  to  lower  the  total  circuit  inductance, 

while  the  addition  of  stators  and  rotors  raises 

the  total  capacitance,  as  well  as  the  ratio  of 
maximum  to  minimum  capacitance. 


234  Radio  Receiver  Fundamentals 


THE  RADIO 


Butterfly  circuits  have  been  applied  specifi- 
cally to  oscillators  for  transmitters,  super- 
heterodyne receivers,  and  heterodyne  frequen- 
cy meters  in  the  100-1000-Mc.  frequency  range. 

Receiver  The  types  of  resonant  circuits  de- 
Circuits  scribed  in  the  previous  paragraphs 
have  largely  replaced  conventional 
coil-capacitor  circuits  in  the  range  above  100 
Me.  Tuned  short  lines  and  butterfly  circuits 
are  used  in  the  range  from  about  100  Me.  to 
perhaps  3500  Me.,  and  above  about  3500  Me. 
resonant  cavities  are  used  almost  exclusively. 
The  resonant  cavity  is  also  quite  generally 
employed  in  the  2000-Mc.  to  3500-Mc.  range. 

In  a properly  designed  receiver,  thermal 
agitation  in  the  first  tuned  circuit  is  amplified 
by  subsequent  tubes  and  predominates  in  the 
output.  For  good  signal-to-set-noise  ratio, 
therefore,  one  must  strive  for  a high-gain  low- 
noise  r-f  stage.  Hiss  can  be  held  down  by  giv- 
ing careful  attention  to  this  point.  A mixer 
has  about  0.3  of  the  gain  of  an  r-f  tube  of  the 
same  type;  so  it  is  advisable  to  precede  a 
mixer  by  an  efficient  r-f  stage.  It  is  also  of 
some  value  to  have  good  t-f  selectivity  before 
the  first  detector  in  order  to  reduce  noises 
produced  by  beating  noise  at  one  frequency 
against  noise  at  another,  to  produce  noise  at 
the  intermediate  frequency  in  a superheter- 
odyne. 

The  frequency  limit  of  a tube  is  reached 
when  the  shortest  possible  external  connec- 
tions are  used  as  the  tuned  circuit,  except  for 
abnormal  types  of  oscillation.  Wires  or  size- 
able components  are  often  best  considered  as 
sections  of  transmission  lines  rather  than  as 
simple  resistances,  capacitances,  or  induct- 
ances. 

So  long  as  small  ttiodes  and  pentodes  will 
operate  normally,  they  are  generally  preferred 
as  v-h-f  tubes  over  other  receiving  methods 
that  have  been  devised.  However,  the  input 
capacitance,  input  conductance,  and  transit 
time  of  these  tubes  limit  the  upper  frequency 
at  which  they  may  be  operated.  The  input  re- 
sistance, which  drops  to  a low  value  at  very 
short  wave-lengths,  limits  the  stage  gain  and 
broadens  the  tuning. 

V-H-F  The  first  tube  in  a v-h-f  receiver  is 
Tubes  most  important  in  raising  the  signal 
above  the  noise  generated  in  succes- 
sive stages,  for  which  reason  small  v-h-f  types 
are  definitely  preferred. 

Tubes  employing  the  conventional  grid-con- 
trolled  and  diode  rectifier  principles  have  been 
modernized,  through  various  expedients,  for 
operation  at  frequencies  as  high,  in  some  new 
types,  as  4000  Me.  Beyond  that  frequency, 
electron  transit  time  becomes  the  limiting  fac- 


tor and  new  principles  must  be  enlisted.  In 

general,  the  improvements  embodied  in  exist- 
ing tubes  have  consisted  of  (1)  reducing  elec- 
trode spacing  to  cut  down  electron  transit 
time,  (2)  reducing  electrode  areas  to  decrease 
interelectrode  capacitances,  and  (3)  shorten- 
ing of  electrode  leads  either  by  mounting  the 
electrode  assembly  close  to  the  tube  base  or 
by  bringing  the  leads  out  directly  through  the 
glass  envelope  at  nearby  points.  Through  re- 
duction of  lead  inductance  and  intetelecttode 
capacitances,  input  and  output  resonant  fre- 
quencies due  to  tube  construction  have  beet; 
increased  substantially. 

Tubes  embracing  one  or  more  of  the  fea- 
tures just  outlined  include  the  later  loctal 
types,  high-frequency  acorns,  button-base 
types,  and  the  lighthouse  types.  Type  6j4 
button-base  ttiode  will  reach  500  Me.  Type 
6F4  acorn  triode  is  recommended  for  use  up  to 
1200  Me.  Type  1A3  button-base  diode  has  a 
resonant  frequency  of  1000  Me.,  while  type 
9005  acorn  diode  resonates  at  1500  Me.  Light- 
house type  2C40  can  be  used  at  frequencies 
up  to  3500  Me.  as  an  oscillator. 

Crystal  More  than  two  decades  have 
Rectifiers  passed  since  the  crystal  (mineral) 
rectifier  enjoyed  widespread  use 
in  radio  receivers.  Low-priced  tubes  complete- 
ly supplanted  the  fragile  and  relatively  insen- 
sitive crystal  detector,  although  it  did  con- 
tinue for  a few  years  as  a simple  meter  recti- 
fier in  absorption  wavemeters  after  its  demise 
as  a receiver  component. 

Today,  the  crystal  detector  is  of  new  im- 
portance in  microwave  communication.  It  is 
being  employed  as  a detector  and  as  a mixer 
in  receivers  and  test  instruments  used  at  ex- 
tremely high  radio  frequencies.  At  some  of 
the  frequencies  employed  in  microwave  opera- 
tions, the  crystal  rectifier  is  the  only  satis- 
factory detector  or  mixer.  The  chief  advan- 
tages of  the  crystal  rectifier  are  very  low  ca- 
pacitance, relative  freedom  from  transit-time 
difficulties,  and  its  two-terminal  nature.  No 
batteries  or  a-c  power  supply  are  required  for 
its  operation. 

The  crystal  detector  consists  essentially  of 
a small  piece  of  silicon  or  germanium  mount- 
ed in  a base  of  low-melting-point  alloy  and 
contacted  by  means  of  a thin,  springy  feeler 
wire  known  as  the  cat  whisker.  This  arrange- 
ment is  shown  in  figure  38A. 

The  complex  physics  of  crystal  rectification 
is  beyond  the  scope  of  this  discussion.  It  is 
sufficient  to  state  that  current  flows  from  sev- 
eral hundred  to  several  thousand  times  mote 
readily  in  one  direction  through  the  contact  of 
cat  whisker  and  crystal  than  in  the  opposite 
direction.  Consequently,  an  alternating  current 
(including  one  of  microwave  frequency)  will 


HANDBOOK 


Receiver  Adjustment  235 


Figure  38 

1N23  MICROWAVE-TYPE 
CRYSTAL  DIODE 


A small  silicon  crystal  is  attached  to 
the  base  connector  and  a fine  **cat- 
whisker**  wire  is  set  to  the  most  sensi- 
tive spot  on  the  crystal.  After  adjust- 
ment the  ceramic  shell  is  filled  with 
compound  to  hold  the  contact  wire  in 
position.  Crystals  of  this  type  are  used 
to  over  30,000  me. 

be  rectified  by  the  crystal  detector.  The  load, 
through  which  the  rectified  currents  flow,  may 
be  connected  in  series  or  shunt  with  the  crys- 
tal, although  the  former  connection  is  most 
generally  employed. 

The  basic  arrangement  of  a modern  fixed 
crystal  detectot  developed  during  World  War  II 
for  microwave  work,  particularly  radar,  is 
shown  in  figure  38B.  Once  the  cat  whisker  of 
this  unit  is  set  at  the  factory  to  the  most  sen- 
sitive spot  on  the  surface  of  the  silicon  crys- 
tal and  its  pressure  is  adjusted,  a -filler  com- 
pound is  injected  through  the  filling  hole  to 
hold  the  cat  whisker  permanently  in  position. 


12-11  Receiver  Adjustment 


A simple  regenerative  receiver  requires 
little  adjustment  other  than  that  necessary 
to  insure  correct  tuning  and  smooth  regenera- 
tion over  some  desired  range.  Receivers  of 
the  tuned  radio-frequency  type  and  superheter- 
odynes require  precise  alignment  to  obtain  the 
highest  possible  degree  of  selectivity  and 
sensitivity. 

Good  results  can  be  obtained  from  a receiv- 
er only  when  it  is  properly  aligned  and  ad- 
justed. The  most  practical  technique  for  mak- 
ing these  adjustments  is  given  below. 

Instruments  A very  small  number  of  instru- 
ments will  suffice  to  check  and 
align  a communications  receiver,  the  most  im- 
portant of  these  testing  units  being  a modu- 
lated oscillator  and  a d-c  and  a-c  voltmeter. 
The  meters  are  essential  in  checking  the  volt- 
age applied  at  each  circuit  point  from  the  pow- 
er supply.  If  the  a-c  voltmeter  is  of  the  oxide- 
rectifier  type,  it  can  be  used,  in  addition,  as 
an  output  meter  when  connected  across  the 


receiver  output  when  tuning  to  a modulated 
signal.  If  the  signal  is  a steady  tone,  such  as 
from  a test  oscillator,  the  output  meter  will 
indicate  the  value  of  the  detected  signal.  In 
this  manner,  alignment  results  may  be  visually 
noted  on  the  meter. 

T-R-F  Receiver  Alignment  procedure  in  a mul- 
Alignment  tistage  t-r-f  receiver  is  exact- 

ly the  same  as  aligning  a 
single  stage.  If  the  detector  is  regenerative, 
each  preceding  stage  is  successively  aligned 
while  keeping  the  detector  circuit  tuned  to  the 
test  signal,  the  latter  being  a station  signal 
or  one  locally  generated  by  a test  oscillator 
loosely  coupled  to  the  antenna  lead.  During 
these  adjustments,  the  t-f  amplifier  gain  con- 
trol is  adjusted  for  maximum  sensitivity,  as- 
suming that  the  r-f  amplifier  is  stable  and 
does  not  oscillate.  Often  a sensitive  receiver 
can  be  roughly  aligned  by  tuning  for  maximum 
noise  pickup. 

Superheterodyne  Aligning  a superhet  is  a de- 
Alignment  tailed  task  requiring  a great 

amount  of  care  and  patience. 
It  should  never  be  undertaken  without  a thor- 
ough understanding  of  the  involved  job  to  be 
done  and  then  only  when  there  is  abundant 
time  to  devote  to  the  operation.  There  are  no 
short  cuts;  every  circuit  must  be  adjusted  in- 
dividually and  accurately  if  the  receiver  is  to 
give  peak  performance.  The  precision  of  each 
adjustment  is  dependent  upon  the  accuracy 
with  which  the  preceding  one  was  made. 

Superhet  alignment  requites  (1)  a good  sig- 
nal generator  (modulated  oscillator)  covering 
the  radio  and  intermediate  frequencies  and 
equipped  with  an  attenuator;  (2)  the  necessary 
socket  wrenches,  screwdrivers,  or  "neutral- 
izing tools”  to  adjust  the  various  i-f  and  r-f 
trimmer  capacitors;  and  (3)  some  convenient 
type  of  tuning  indicator,  such  as  a copper- 
oxide  or  electronic  voltmeter. 

Throughout  the  alignment  process,  unless 
specifically  stated  otherwise,  the  r-f  gain  con- 
trol must  be  set  for  maximum  output,  the  beat 
oscillator  switched  off,  and  the  a.v.c.  turned 
off  or  shorted  out.  When  the  signal  output  of 
the  receiver  is  excessive,  either  the  attenuator 
or  the  a-f  gain  control  may  be  turned  down,  but 
never  the  r-f  gain  control. 

I-F  Alignment  After  the  receiver  has  been 
given  a rigid  electrical  and 
mechanical  inspection,  and  any  faults  which 
may  have  been  found  in  wiring  or  the  selec- 
tion and  assembly  of  parts  corrected,  the  i-f 
amplifier  may  be  aligned  as  the  first  step  in 
the  checking  operations. 

With  the  signal  generator  set  to  give  a modu- 
lated signal  on  the  frequency  at  which  the  i-f 


236  Radio  Receiver  Fundamentals 


THE  RADIO 


amplifier  is  to  operate,  clip  the  "hot”  output 

lead  from  the  generator  to  the  last  i-f  stage 
through  a small  fixed  capacitor  to  the  control 
grid.  Adjust  both  trimmer  capacitors  in  the 
last  i-f  transformer  (the  one  between  the  last 
i-f  amplifier  and  the  second  detector)  to  reson- 
ance as  indicated  by  maximum  deflection  of 
the  output  meter. 

Each  i-f  stage  is  adjusted  in  the  same  man- 
ner, moving  the  hot  lead,  stage  by  stage,  back 
toward  the  front  end  of  the  receiver  and  back- 
ing off  the  attenuator  as  the  signal  strength 
increases  in  each  new  position.  The  last  ad- 
justment will  be  made  to  the  first  i-f  trans- 
former, with  the  hot  signal  generator  lead  con- 
nected to  the  control  grid  of  the  mixer.  Oc- 
casionally it  is  necessary  to  disconnect  the 
mixer  grid  lead  from  the  coil,  grounding  it 
through  a 1,000-  or  5,000-ohm  resistor,  and 
coupling  the  signal  generator  through  a small 
capacitor  to  the  grid. 

When  the  last  i-f  adjustment  has  been  com- 
pleted, it  is  good  practice  to  go  back  through 
the  i-f  channel,  re-peaking  all  of  the  trans- 
formers. It  is  imperative  that  this  recheck  be 
made  in  sets  which  do  not  include  a crystal 
filter,  and  where  the  simple  alignment  of  the 
i-f  amplifier  to  the  generator  is  final. 

I-F  with  There  are  several  ways  of  align- 

Crystal  Filter  ing  an  i-f  channel  which  con- 
tains a crystal-filter  circuit. 
However,  the  following  method  is  one  which 
has  been  found  to  give  satisfactory  results  in 
every  case:  An  unmodulated  signi  generator 
capable  of  tuning  to  the  frequency  of  the  filter 
crystal  in  the  receiver  is  coupled  to  the  grid  of 
the  stage  which  precedes  the  crystal  filter  in 
the  receiver.  Then,  with  the  crystal  filter 
switched  in,  the  signal  generator  is  tuned 
slowly  to  find  the  frequency  where  the  crystal 
peaks.  The  receiver  "S”  meter  may  be  used 
as  the  indicator,  and  the  sound  heard  from  the 
loudspeaker  will  be  of  assistance  in  finding 
the  point.  When  the  frequency  at  which  the 
crystal  peaks  has  been  found,  all  the  i-f  trans- 
formers in  the  receiver  should  be  touched  up 
to  peak  at  that  frequency. 

B-F-0  Adlustment  Adjusting  the  beat  oscil- 
latoron  a receiver  that  has 
no  front  panel  adjustment  is  relatively  simple. 
It  is  only  necessary  to  tune  the  receiver  to 
resonance  with  any  signal,  as  indicated  by 
the  tuning  indicator,  and  then  turn  on  the  b.f.o. 
and  set  its  trimmer  (or  trimmers)  to  produce 
the  desired  beat  note.  Setting  the  beat  oscil- 
lator in  this  way  will  result  in  the  beat  note 
being  stronger  on  one  **side**  of  the  signal 
than  on  the  other,  which  is  what  is  desired 
for  c-w  reception.  The  b.f.o.  should  not  be  set 
to  zero  heat  when  the  receiver  is  tuned  to 


Figure  39 

THE  Q-MULTIPLIER 

The  loss  resistance  of  a high-Q  circuit 
is  neutralized  by  regeneration  in  a 
simple  feedback  amplifier.  A highly 
selective  passband  is  produced  which 
is  coupled  to  the  i-f  circuit  of  the 
receiver. 


resonance  with  the  signal,  as  this  will  cause 
an  equally  strong  beat  to  be  obtained  on  both 
sides  of  resonance. 

Front-End  Alignment  of  the  front  end  of  a 
Alignment  home-constructed  receiver  is  a 
relatively  simple  process,  con- 
sisting of  first  getting  the  oscillator  to  cover 
the  desired  frequency  range  and  then  of  peak- 
ing the  various  r-f  circuits  for  maximum  gain. 
However,  if  the  frequency  range  covered  by 
the  receiver  is  very  wide  a fair  amount  of  cut 
and  try  will  be  requited  to  obtain  satisfactory 
tracking  between  the  t-f  circuits  and  the  oscil- 
lator. Manufactured  communications  receivers 
should  always  be  tuned  in  accordance  with 
the  instructions  given  in  the  maintenance  man- 
ual for  the  receiver. 


12-12  Receiving  Accessories 

The  Q-Multiplier  The  selectivity  of  a receiver 
may  be  increased  by  raising 
the  Q of  the  tuned  circuits  of  the  i-f  strip.  A 
simple  way  to  accomplish  this  is  to  add  a con- 
trolled amount  of  positive  feedback  to  a tuned 
circuit,  thus  increasing  its  Q.  This  is  done  in 
the  (^-multiplier,  whose  basic  circuit  is  shown 
in  figure  39.  The  circuit  L-C1-C2  is  tuned  to 


HANDBOOK 


Receiving  Accessories  237 


465  KC 
FREQUENCY 


- B 200-300  V.  ■=■  6.3  V. 


V.^~GRArBUffNE  Vt  ZHOKZ  {0.6-6.0MH) 
\-Z=-GRAYBURNE  "LOOPSTiCK"  COIL 


Figure  40 

q-multiplier  null  circuit 

The  addition  of  a second  triode  permits 
the  Q*Mu/f/p//er  to  be  used  for  nulling 
out  an  unwanted  hetrodyne. 


the  intermediate  frequency,  and  the  loss  re- 
sistance of  the  circuit  is  neutralized  by  the 
positive  feedback  circuit  composed  of  C3  and 
the  vacuum  tube.  Too  great  a degree  of  positive 
feedback  will  cause  the  circuit  to  break  into 
oscillation. 

At  the  resonant  frequency,  the  impedance  of 
the  tuned  circuit  is  very  high,  and  when  shunted 
across  an  i-f  stage  will  have  little  effect  upon 
the  signal.  At  frequencies  removed  from  reso- 
nance, the  impedance  of  the  circuit  is  low,  re- 
sulting in  high  attenuation  of  the  i-f  signal. 
The  resonant  frequency  of  the  Q-multipliec  may 
by  varied  by  changing  the  value  of  one  of  the 
components  in  the  tuned  circuit. 

The  Q-multiplier  may  also  be  used  to  ”null’* 
a signal  by  employing  negative  feedback  to 
control  the  plate  resistance  of  an  auxiliary 
amplifier  stage  as  shown  in  figure  40.  Since  the 
grid-cathode  phase  shift  through  the  Q-multiplier 
is  zero,  the  plate  resistance  of  a second  tube 
may  be  readily  controlled  by  placing  it  across 
the  Q-multiplier.  At  resonance,  the  high  nega- 
tive feedback  drops  the  plate  resistance  of 
V2,  shunting  the  i-f  circuit.  Off  resonance,  the 
feedback  is  reduced  and  the  plate  resistance 
of  V2  rises,  reducing  the  amount  of  signal  at- 
tenuation in  the  i-f  strip.  A circuit  combining 
both  the  "peak”  and  "null”  features  is  shown 
in  figure  41. 

The  Product  Detector  A version  of  the  common 
mixer  or  converter  stage 


Figure  41 

SCHEMATIC  OF  A 455KC 
Q-MULTIPLIER 

Coil  Li  is  required  to  tune  out  the 
reactance  of  the  coaxial  line.  It  is  ad- 
justed for  maximum  signal  response. 
LI  may  be  om»tted  if  the  Q-mulfiplier 
is  conneefed  to  the  receiver  with  o 
short  length  of  wire,  and  the  i-f  trans- 
former within  the  receiver  is  retuned. 


may  be  used  as  a second  detector  in  a receiver 
in  place  of  the  usual  diode  detector.  The  diode 
is  an  envelope  detector  (section  12-1)  and  de- 
velops a d-c  output  voltage  from  a single  r-f 
signal,  and  audio  "beats”  from  two  or  more 
input  signals.  A product  detector  (figure  42) 
requires  that  a local  carrier  voltage  be  present 
in  order  to  produce  an  audio  output  signal. 


Figure  42 
THE  PRODUCT 
DETECTOR 

Audio  output  sf'gna/  is 
developed  only  when 
local  oscillator  is  on. 


238 


Radio  Receiver  Fundamentals 


Figure  43 

PENTAGRID  MIXER 
USED  AS  PRODUCT 
DETECTOR 


Such  a detector  is  useful  for  single  sideband 
work,  since  the  inter-modulation  distortion  is 
extremely  low. 

A pentagrid  product  detector  is  shown  in 
figure  43-  The  incoming  signal  is  applied  to 
grid  3 of  the  mixer  tube,  and  the  local  oscillator 
is  injected  on  grid  1.  Grid  bias  is  adjusted  for 
operation  ovet  the  linear  portion  of  the  tube 
characteristic  curve.  When  grid  1 injection 
is  removed,  the  audio  output  from  an  unmodu- 
lated signal  applied  to  grid  3 should  be  reduced 
approximately  30  to  40  db  below  normal  de- 
tection level.  When  the  frequency  of  the  local 
oscillator  is  synchronized  with  the  incoming 
carrier,  amplitude  modulated  signals  may  be 
received  by  exalted  carrier  reception,  wherein 
the  local  carrier  substitutes  for  the  transmitted 
carrier  of  the  a-m  signal. 

Three  triodes  may  be  used  as  a product 
detector  (figure  44).  Triodes  VI  and  V2  act 
as  cathode  followers,  delivering  the  sideband 
signal  and  the  local  oscillator  signal  to  a 
grounded  grid  triode  (V3)  which  functions  as 
the  mixer  stage.  A third  version  of  the  product 
detector  is  illustrated  in  figure  45.  A twin 
triode  tube  is  used.  Section  VI  functions  as 
a cathode  follower  amplifier.  Section  V2  is  a 


Figure  44 
TRIPLE-TRIODE 
PRODUCT  DETECTOR 

V7  and  V2  act  as  cathode  follow- 
ers, del ivering  sideband  signal 
and  local  oscillator  signal  to 
grounded  grid  triode  mixer  (V3). 


PRODUCT  DETECTOR 


**plate”  detector,  the  cathode  of  which  is 
common  with  the  cathode  follower  amplifier. 
The  local  oscillator  signal  is  injected  into  the 
grid  circuit  of  tube  V2. 


Figure  46 

EXPLODED  VIEW  OF  COLLINS 
MECHANICAL  FILTER 


CHAPTER  THIRTEEN 


Generation  of 
Radio  Frequency  Energy 


A radio  communication  or  broadcast  trans- 
mitter consists  of  a source  of  radio  frequency 
power,  or  carrier!  a system  for  modulating  the 
carrier  whereby  voice  or  telegraph  keying  or 
other  modulation  is  superimposed  upon  it;  and 
an  antenna  system,  including  feed  line,  for 
radiating  the  intelligence-carrying  radio  fre- 
quency power.  The  power  supply  employed  to 
convert  primary  power  to  the  various  voltages 
required  by  the  r-f  and  modulator  portions  of 
the  transmitter  may  also  be  considered  part 
of  the  transmitter. 

Voice  modulation  usually  is  accomplished 
by  varying  either  the  amplitude  or  the  frequen- 
cy of  the  radio  frequency  carrier  in  accordance 
with  the  components  of  intelligence  to  be 
transmitted. 

Radiotelegraph  modulation  (keying)  normal- 
ly is  accomplished  either  by  interrupting, 
shifting  the  frequency  of,  or  superimposing  an 
audio  tone  on  the  radio-frequency  carrier  in 
accordance  with  the  dots  and  dashes  to  be 
transmitted. 

The  complexity  of  the  radio- frequency  gen- 
erating portion  of  the  transmitter  is  dependent 
upon  the  power,  order  of  stability,  and  frequen- 
cy desired.  An  oscillator  feeding  an  antenna 
directly  is  the  simplest  form  of  radio-frequency 
generator.  A modern  high-frequency  transmitter, 
on  the  other  hand,  is  a very  complex  generator. 
Such  an  equipment  usually  comprises  a very 


stable  crystal-controlled  or  self-controlled 
oscillator  to  stabilize  the  output  frequency,  a 
series  of  frequency  multipliers,  one  or  more 
amplifier  stages  to  increase  the  power  up  to 
the  level  which  is  desired  for  feeding  the  an- 
tenna system,  and  a filter  system  for  keeping 
the  harmonic  energy  generated  in  the  trans- 
mitter from  being  fed  to  the  antenna  system. 


13-1  Self-Controlled 

Oscillators 

In  Chapter  Four,  it  was  explained  that  the 
amplifying  properties  of  a tube  having  three 
or  more  elements  give  it  the  ability  to  gener- 
ate an  alternating  current  of  a frequency  de- 
termined by  the  components  associated  with 
it.  A vacuum  tube  operated  in  such  a circuit 
is  called  an  oscillator,  and  its  function  is 
essentially  to  convert  direct  current  into  radio- 
frequency  alternating  current  of  a predeter- 
mined frequency. 

Oscillators  for  controlling  the  frequency  of 
conventional  radio  transmitters  can  be  divided 
into  two  general  classes:  self-controlled  and 

crystal-controlled. 

There  are  a great  many  types  of  self-con- 
trolled oscillators,  each  of  which  is  best  suited 


239 


240  Generation  of  R-F  Energy 


THE  RADIO 


@ TUNED-PLATE  UNTUNED  GRID  @ ELECTRON  COUPLED  © COLPITTS  ELECTRON  COUPLED 


© CLAPP  ® CLAPP  ELECTRON  COUPLED 


Figure  1 

COMMON  TYPES  OF  SELF-EXCITED  OSCILLATORS 

Fixed  capacitor  values  are  typical,  but  will  vary  somewhat  with  the  application. 
In  the  Clapp  oscillator  circuits  (G)  and  (H),  capacitors  Cj  and  C2  should  have  a 
reactance  of  50  to  100  ohms  at  the  operating  frequency  of  the  osciliator.  Tuning  of 
th^ese  two  oscillators  is  occomplished  by  capacitor  C.  In  the  circuits  of  (E),  (F),  and 
(H),  tuning  of  the  tank  circuit  in  the  plate  of  the  oscillator  tube  will  have  relatively 
small  effect  on  the  frequency  of  oscillation.  The  plate  tank  circuit  also  may,  if  de- 
sired,  be  tuned  to  a harmonic  of  the  osciUation  frequency,  or  o broadly  resonant 
circuit  may  be  used  in  this  circuit  position. 


to  a particular  application.  They  can  further 
be  subdivided  into  the  classifications  of:  neg- 
ative-grid oscillators,  electron-orbit  oscilla- 
tors, negative-resistance  oscillators,  velocity 
modulation  oscillatots,  and  magnetron  oscil- 
lators. 

Negative-Grid  A negative-grid  oscillator  is 

Oscillators  essentially  a vacuum-tube  am- 

plifier with  a sufficient  por- 
tion of  the  output  energy  coupled  back  into  the 
input  circuit  to  sustain  oscillation.  The  con- 


trol grid  is  biased  negatively  with  respect  to 
the  cathode.  Common  types  of  negative-grid 
oscillators  are  diagrammed  in  figure  1. 

The  Hartley  Illustrated  in  figure  1 (A)  is  the 
oscillator  circuit  which  finds  the 
most  general  application  at  the  present  time; 
this  circuit  is  commonly  called  the  Hartley. 
The  operation  of  this  oscillator  will  be  de- 
scribed as  an  index  to  the  operation  of  all 
negative-grid  oscillators;  the  only  teal  differ- 


HANDBOOK 


Oscillators  241 


ence  between  the  various  circuits  is  the  man- 
ner in  which  energy  for  excitation  is  coupled 
from  the  plate  to  the  grid  circuit. 

When  plate  voltage  is  applied  to  the  Hartley 
oscillator  shown  at  (A),  the  sudden  flow  of 
plate  current  accompanying  the  application  of 
plate  voltage  will  cause  an  electro-magnetic 
field  to  be  set  up  in  the  vicinity  of  the  coil. 
The  building-up  of  this  field  will  cause  a po- 
tential drdp  to  appear  from  turn-to-turn  along 
the  coil.  Due  to  the  inductive  coupling  be- 
tween the  portion  of  the  coil  in  which  the  plate 
current  is  flowing  and  the  grid  portion,  a po- 
tential will  be  induced  in  the  grid  portion. 

Since  the  cathode  tap  is  between  the  grid 
and  plate  ends  of  the  coil,  the  induced  grid 
voltage  acts  in  such  a manner  as  to  increase 
further  the  plate  current  to  the  tube.  This  ac- 
tion will  continue  for  a short  period  of  time 
determined  by  the  inductance  and  capacitance 
of  the  tuned  circuit,  until  the  flywheel  effect 
of  the  tuned  circuit  causes  this  action  to  come 
to  a maximum  and  then  to  reverse  itself.  The 
plate  current  then  decreases,  the  magnetic 
field  around  the  coil  also  decreasing,  until  a 
minimum  is  reached,  when  the  action  starts 
again  in  the  original  direction  and  at  a greater 
amplitude  than  before.  The  amplitude  of  these 
oscillations,  the  frequency  of  which  is  de- 
termined by  the  coil-capacitor  circuit,  will  in- 
crease in  a very  short  period  of  time  to  a limit 
determined  by  the  plate  voltage  of  the  oscil- 
lator tube. 

The  Colpitts  Figure  1 (B)  shows  a version  of 
the  Colpitts  oscillator.  It  can 
be  seen  that  this  is  essentially  the  same  cir- 
cuit as  the  Hartley  except  that  the  ratio  of  a 
pair  of  capacitances  in  series  determines  the 
effective  cathode  tap,  instead  of  actually  us- 
ing a tap  on  the  tank  coil.  Also,  the  net  ca- 
pacitance of  these  two  capacitors  comprises 
the  tank  capacitance  of  the  tuned  circuit.  This 
oscillator  circuit  is  somewhat  less  suscep- 
tible to  parasitic  (spurious)  oscillations  than 
the  Hartley. 

For  best  operation  of  the  Hartley  and  Col- 
pitts oscillators,  the  voltage  from  grid  to  cath- 
ode, determined  by  the  tap  on  the  coil  or  the 
setting  of  the  two  capacitors,  normally  should 
be  from  1/3  to  1/5  that  appearing  between 
plate  and  cathode. 

The  T.P.T.G.  The  tuned-plate  tuned-grid  os- 
cillator illustrated  at  (C)  has 
a tank  circuit  in  both  the  plate  and  grid  cir- 
cuits. The  feedback  of  energy  from  the  plate 
to  the  grid  circuits  is  accomplished  by  the 
plate-to-grid  intet-electrode  capacitance  with- 
in the  tube.  The  necessary  phase  reversal  in 
feedback  voltage  is  provided  by  tuning  the 
grid  tank  capacitor  to  the  low  side  of  the  de- 


sired frequency  and  the  plate  capacitor  to  the 
high  side.  A broadly  resonant  coil  may  be  sub- 
stituted for  the  grid  tank  to  form  the  T.N.T, 
oscillator  shown  at  (D). 

Electron-Coupled  In  any  of  the  oscillator  cir- 
Oscillotors  cuits  just  described  it  is 

possible  to  take  energy  from 
the  oscillator  circuit  by  coupling  an  external 
load  to  the  tank  circuit.  Since  the  tank  circuit 
determines  the  frequency  of  oscillation  of  the 
tube,  any  variations  in  the  conditions  of  the 
external  circuit  will  be  coupled  back  into  the 
frequency  determining  portion  of  the  oscillator. 
These  variations  will  result  in  frequency  in- 
stability. 

The  frequency  determining  portion  of  an 
oscillator  may  be  coupled  to  the  load  circuit 
only  by  an  electron  stream,  as  illustrated  in 
(E)  and  (F)  of  figure  1.  When  it  is  considered 
that  the  screen  of  the  tube  acts  as  the  plate 
to  the  oscillator  circuit,  the  plate  merely  act- 
ing as  a coupler  to  the  load,  then  the  sim- 
ilarity between  the  cathode-grid-screen  circuit 
of  these  oscillators  and  the  cathode-grid-plate 
circuits  of  the  corresponding  prototype  can  be 
seen. 

The  electron-coupled  oscillator  has  good 
stability  with  respect  to  load  and  voltage  var- 
iation. Load  variations  have  a relatively  small 
effect  on  the  frequency,  since  the  only  cou- 
pling between  the  oscillating  circuit  and  the 
load  is  through  the  electron  stream  flowing 
through  the  other  elements  to  the  plate.  The 
plate  is  electrostatically  shielded  from  the 
oscillating  portion  by  the  bypassed  screen. 

The  stability  of  the  e.c.o.  with  respect  to 
variations  in  supply  voltages  is  explained  as 
follows:  The  frequency  will  shift  in  one  direc- 
tion with  an  increase  in  screen  voltage,  while 
an  increase  in  plate  voltage  will  cause  it  to 
shift  in  the  other  direction.  By  a proper  pro- 
portioning of  the  resistors  that  comprise  the 
voltage  divider  supplying  screen  voltage,  it  is 
possible  to  make  the  frequency  of  the  oscil- 
lator substantially  independent  of  supply  volt- 
age variations. 

The  Clapp  A relatively  new  type  of  oscillator 
Oscillator  circuit  which  is  capable  of  giving 
excellent  frequency  stability  is 
illustrated  in  figure  IG.  Comparison  between 
the  more  standard  circuits  of  figure  lA  through 
IF  and  the  Clapp  oscillator  circuits  of  figures 
IG  and  IH  will  immediately  show  one  marked 
difference:  the  tuned  circuit  which  controls 
the  operating  frequency  in  the  Clapp  oscillator 
is  series  resonant,  while  in  all  the  more  stand- 
ard oscillator  circuits  the  frequency  control- 
ling circuit  is  parallel  resonant.  Also,  the 
capacitors  C,  and  C2  are  relatively  large  in 
terms  of  the  usual  values  for  a Colpitts  oscil- 


242  Generation  of  R-F  Energy 


THE  RADIO 


lator.  In  fact,  the  value  of  capacitors  Cj  and 
Cj  will  be  in  the  vicinity  of  0.001  fiid.  to 
0.0025  fifd.  for  an  oscillator  which  is  to  be 
operated  in  the  1.8-Mc.  band. 

The  Clapp  oscillator  operates  in  the  follow- 
ing manner;  at  the  resonant  frequency  of  the 
oscillator  runed  circuit  (L,  C)  the  impedance 
of  this  circuit  is  at  minimum  (since  it  oper- 
ates in  series  resonance)  and  maximum  cur- 
rent flows  through  it.  Note  however,  that  Cj 
and  Cj  also  are  included  within  the  current 
path  for  the  series  resonant  circuit,  so  that  at 
the  frequency  of  resonance  an  appreciable 
voltage  drop  appears  across  these  capacitors. 
The  voltage  drop  appearing  across  C,  is  ap- 
plied to  the  grid  of  the  oscillator  tube  as  ex- 
citation, while  the  amplified  output  of  the 
oscillator  tube  appears  across  Cj  as  the  driv- 
ing power  to  keep  the  circuit  in  oscillation. 

Capacitors  C,  and  Cj  should  be  made  as 
large  in  value  as  possible,  while  still  permit- 
ting the  circuit  to  oscillate  over  the  full  tun- 
ing range  of  C.  The  larger  these  capacitors 
are  made,  the  smaller  will  be  the  coupling  be- 
tween the  oscillating  circuit  and  the  tube,  and 
consequently  the  better  will  be  oscillator  sta- 
bility with  respect  to  tube  variations.  High  G„ 
tubes  such  as  the  6AC7,  6AG7,  and  6CB6  will 
permit  the  use  of  larger  values  of  capacitance 
at  C,  and  Cj  than  will  more  conventional  tubes 
such  as  the  6SJ7,  6V6,  and  such  types.  In  gen- 
eral it  may  be  said  that  the  reactance  of  ca- 
pacitors C,  and  C}  should  be  on  the  order  of 
40  to  120  ohms  at  the  operating  frequency  of 
the  oscillator— with  the  lower  values  of  re- 
actance going  with  high-G„,  tubes  and  the 
higher  values  being  necessary  to  permit  oscil- 
lation with  tubes  having  G^  in  the  range  of 
2000  micromhos  such  as  the  6SJ7. 

It  will  be  found  that  the  Clapp  oscillator 
will  have  a tendency  to  vary  in  power  output 
over  the  frequency  range  of  tuning  capacitor 
C.  The  output  will  be  greatest  where  C is  at 
its  largest  setting,  and  will  tend  to  fall  off 
with  C at  minimum  capacitance.  In  fact,  if 
capacitors  C,  and  Cj  have  too  large  a value 
the  circuit  will  stop  oscillation  near  the  mini- 
mum capacitance  setting  of  C.  Hence  it  will 
be  necessary  to  use  a slightly  smaller  value 
of  capacitance  at  C,  and  Cj  (to  provide  an  in- 
crease in  the  capacitive  reactance  at  this 
point),  or  else  the  frequency  range  of  the  oscil- 
lator must  be  restricted  by  paralleling  a fixed 
capacitor  across  C so  that  its  effective  capaci- 
tance at  minimum  setting  will  be  increased  to 
a value  which  will  sustain  oscillation. 

In  the  triode  Clapp  oscillator,  such  as  shown 
at  figure  IG,  output  voltage  for  excitation  of 
an  amplifier,  doubler,  or  isolation  stage  nor- 
mally is  taken  from  the  cathode  of  the  oscil- 
lator tube  by  capacitive  coupling  to  the  grid 
of  the  next  tube.  However,  where  greater  iso- 


lation of  succeeding  stages  from  the  oscillat- 
ing circuit  is  desired,  the  electron-coupled 
Clapp  oscillator  diagrammed  in  figure  IH  may 
be  used.  Output  then  may  be  taken  from  the 
plate  circuit  of  the  tube  by  capacitive  coupling 
with  either  a tuned  circuit,  as  shown,  or  with 
an  t-f  choke  or  a broadly  resonant  circuit  in 
the  plate  return.  Alternatively,  energy  may  be 
coupled  from  the  output  circuit  Lj-C,  by  link 
coupling.  The  considerations  with  regard  to 
C„  Cj,  and  the  gtid  tuned  circuit  are  the  same 
as  for  the  triode  oscillator  arrangement  of 
figure  IG. 

Negative  Resist-  Negative-resistance  oscil- 
ance  Oscillators  lators  often  are  used  when 
unusually  high  frequency 
stability  is  desired,  as  in  a frequency  meter. 
The  dynatron  of  a few  years  ago  and  the  newer 
transitron  are  examples  of  oscillator  circuits 
which  make  use  of  the  negative  resistance 
characteristic  between  different  elements  in 
some  multi-grid  tubes. 

In  the  dynatron,  the  negative  resistance  is  a 
consequence  of  secondary  emission  of  elec- 
trons from  the  plate  of  a tetrode  tube.  By  a 
proper  proportioning  of  the  electrode  voltage, 
an  increase  in  screen  voltage  will  cause  a 
decrease  in  screen  current,  since  the  increased 
screen  voltage  will  cause  the  screen  to  attract 
a larger  number  of  the  secondary  electrons 
emitted  by  the  plate.  Since  the  net  screen  cur- 
rent flowing  from  the  screen  supply  will  be 
decreased  by  an  increase  in  screen  voltage, 
it  is  said  that  the  screen  circuit  presents  a 
negative  resistance. 

If  any  type  of  tuned  circuit,  or  even  a re- 
sistance-capacitance circuit,  is  connected  in 
series  with  the  screen,  the  arrangement  will 
oscillate— provided,  of  course,  that  the  external 
circuit  impedance  is  greater  than  the  negative 
resistance.  A negative  resistance  effect  simi- 
lar to  the  dynatron  is  obtained  in  the  transitron 
circuit,  which  uses  a pentode  with  the  suppres- 
sor coupled  to  the  screen.  The  negative  re- 
sistance in  this  case  is  obtained  from  a com- 
bination of  secondary  emission  and  inter-elec- 
trode coupling,  and  is  considerably  more  stable 
than  that  obtained  from  uncontrolled  secondary 
emission  alone  in  the  dynatron.  A representa- 
tive transitron  oscillator  circuit  is  shown  in 
figure  2. 

The  chief  distinction  between  a conven- 
tional negative  grid  oscillator  and  a negative 
resistance  oscillator  is  that  in  the  former  the 
tank  circuit  must  act  as  a phase  inverter  in 
order  to  permit  the  amplification  of  the  tube 
to  act  as  a negative  resistance,  while  in  the 
latter  the  tube  acts  as  its  own  phase  inverter. 
Thus  a negative  resistance  oscillator  requires 
only  an  untapped  coil  and  a single  capacitor 


HANDBOOK 


Oscillators  243 


6SK7 


@ TRANSITRON  OSCILLATOR 


6SN7  OR  6J6 


(i)  CATHODE  COUPLED  OSCILLATOR 


Figure  2 

TWO-TERMINAL  OSCILLATOR  CIRCUITS 

Bof/i  circuits  may  be  used  for  an  audio 
oscillator  or  for  frequencies  into  the 
y-h-f  range  simply  by  placing  a tank  cir- 
cuit tuned  to  the  proper  frequency  where 
indicated  on  the  drawing.  Recommen</e</ 
values  for  the  components  ore  given  be- 
low for  both  oscillators. 


TRANSITION  OSCILLATOR 

Cj>~0.01>/Afd.  mica  for  r.f.  10-/Afd.  elect,  for  o.f. 
Cj— 0.00005- A^-fd.  mica  for  r.f.  0.1«/xfd.  paper  for 
o.f. 

C3— 0.003- /Ufd.  mica  for  r.f.  0.5-A^fd.  paper  for 
o.f. 

C4— O.OI-Aifd.  mice  for  r.f.  8-/xfd.  elect,  for  o.f. 

Rj^220K  H-watt  corbon 

R2— >1800  ohms  H-wott  carbon 

Rj^— 22k  2-wott  carbon 

R4— ^22K  2-watt  carbon 


CATHODE-COUPLED  OSCILLATOR 
Cj^— 0.0000S-/Afd.  mica  for  r.f.  0.1-/i-fd.  poper 
for  oudio 

Cj— -0,003- /xfd.  mico  for  r.f.  8-/Afd.  elect,  for 
oudio 

Rj— -47K  H-wott  carbon 
Rj— IK  1-wott  carbon 


as  the  frequency  determining  tank  circuit,  and 
is  classed  as  a two  terminal  oscillator.  In  fact, 
the  time  constant  of  an  R/C  circuit  may  be 
used  as  the  frequency  determining  element  and 
such  an  oscillator  is  rather  widely  used  as  a 
tunable  audio  frequency  oscillator. 

The  Franklin  The  Franklin  oscillator  makes 
Oscillator  use  of  two  cascaded  tubes  to 

obtain  the  negative-resistance 
effect  (figure  3).  The  tubes  may  be  either  a 
pair  of  triodes,  tetrodes,  or  pentodes,  a dual 
triode,  or  a combination  of  a triode  and  a multi- 
grid tube.  The  chief  advantage  of  this  oscil- 
lator circuit  is  that  the  frequency  determining 
tank  only  has  two  terminals,  and  one  side  of 
the  circuit  is  grounded. 

The  second  tube  acts  as  a phase  inverter  to 
give  an  effect  similar  to  that  obtained  with  the 
dynatron  or  transitron,  except  that  the  effective 
transconductance  is  much  higher.  If  the  tuned 
circuit  is  omitted  or  is  replaced  by  a resistor, 
the  circuit  becomes  a relaxation  oscillator  or 
a multivibrator. 

Oscillator  The  Clapp  oscillator  has  proved 
Stability  to  be  inherently  the  most  stable 
of  all  the  oscillator  circuits  dis- 

cussed  above,  since  minimum  coupling  be- 
tween the  oscillator  tube  and  its  associated 
tuned  circuit  is  possible.  However,  this  in- 


herently good  stability  is  with  respect  to  tube 
variations;  instability  of  the  tuned  circuit  with 
respect  to  vibration  or  temperature  will  of 
course  have  as  much  effect  on  the  frequency  of 
oscillation  as  with  any  other  type  of  oscillator 
circuit.  Solid  mechanical  construction  of  the 
components  of  the  oscillating  circuit,  along 
with  a small  negative- coefficient  compensating 
capacitor  included  as  an  element  of  the  tuned 
circuit,  usually  will  afford  an  adequate  degree 
of  oscillator  stability. 


- B + 


Figure  3 

THE  FRANKLIN  OSCILLATOR  CIRCUIT 

A separate  phase  inverter  tube  is  used  in 
this  oscillator  to  feed  a portion  of  the  output 
back  to  the  input  in  the  proper  phase  to  sus- 
tain  oscillation.  The  values  of  and  C2 
should  be  as  small  as  will  permit  oscillations 
to  be  sustained  over  the  desired  frequency 
range. 


244  Generation  of  R-F  Energy 


THE  RADIO 


V.F.O.  Transmit-  When  used  to  control  the  fre- 

ter  Controls  quency  of  a transmitter  in 

which  there  are  stringent 
limitations  on  frequency  tolerance,  several  pre- 
cautions are  taken  to  ensure  that  a variable 
frequency  oscillator  will  stay  on  frequency. 
The  oscillator  is  fed  from  a voltage  regulated 
power  supply,  uses  a well  designed  and  tem- 
perature compensated  tank  circuit,  is  of  rugged 
mechanical  construction  to  avoid  the  effects 
of  shock  and  vibration,  is  protected  against 
excessive  changes  in  ambient  room  tempera- 
ture, and  is  isolated  from  feedback  or  stray 
coupling  from  other  portions  of  the  transmitter 
by  shielding,  filtering  of  voltage  supply  leads, 
and  incorporation  of  one  or  more  buffer-ampli- 
fier stages.  In  a high  power  transmitter  a small 
amount  of  stray  coupling  from  the  final  ampli- 
fier to  the  oscillator  can  produce  appreciable 
degradation  of  the  oscillator  stability  if  both 
ate  on  the  same  frequency.  Therefore,  the  os- 
cillator usually  is  operated  on  a subharmonic 
of  the  transmitter  output  frequency,  with  one 
or  more  frequency  multipliers  between  the  os- 
cillator and  final  amplifier. 


13-2  Quartz  Crystal 

Oscillators 


Quartz  is  a naturally  occuring  crystal  hav- 
ing a structure  such  that  when  plates  are  cut 
in  certain  definite  relationships  to  the  crystal- 
lographic axes,  these  plates  will  show  the 
piezoelectric  effect— the  plates  will  be  de- 
formed in  the  influence  of  an  electric  field, 
and,  conversely,  when  such  a plate  is  com- 
pressed or  deformed  in  any  way  a potential 
difference  will  appear  upon  its  opposite  sides. 

The  crystal  has  mechanical  resonance,  and 
will  vibrate  at  a very  high  frequency  because 
of  its  stiffness,  the  natural  period  of  vibration 
depending  upon  the  dimensions,  the  method  of 
electrical  excitation,  and  crystallographic 
orientation.  Because  of  the  piezoelectric  pro- 
perties, it  is  possible  to  cut  a quartz  plate 
which,  when  provided  with  suitable  electrodes, 
will  have  the  characteristics  of  a series  reso- 
nant circuit  with  a very  high  L/C  ratio  and 
very  high  Q.  The  Q is  several  times  as  high 
as  can  be  obtained  with  an  inductor-capacitor 
combination  in  conventional  physical  sizes. 
The  equivalent  electrical  circuit  is  shown  in 
figure  4A,  the  resistance  component  simply 
being  an  acknowledgment  of  the  fact  that  the 
Q,  while  high,  does  not  have  an  infinite  value. 

The  shunt  capacitance  of  the  electrodes  and 
associated  wiring  (crystal  holder  and  socket, 
plus  circuit  wiring)  is  represented  by  the  dot- 
ted portion  of  figure  4B.  In  a high  frequency 


Cl 

(small) 


i 


Li  o{ 

(LARG£)^ 

Ri  i 

(small)T 


@ 


f ' 

Li 

3 ' 

o 

C 1 

g "f 

ZCz 

(stray 

SHUNT) 

o 

1 

o 

=pC2 

1 

h ' 

® 

© 

Figure  4 

EQUIVALENT  ELECTRICAL  CIRCUIT  OF 
QUARTZ  PLATE  IN  A HOLDER 

Af  (A)  is  shown  the  equivalent  series-reso- 
nant circuit  of  the  crystal  itself,  at  (B)  is 
shown  how  the  shunt  capacitance  of  the 
holder  electrodes  and  associated  wiring  af- 
fects the  circuit  to  the  combination  circuit 
of  (C)  which  exhibits  both  series  resonance 
and  parallel  resonance  (anti-resonance),  the 
separation  in  frequency  between  the  two 
modes  being  very  small  and  determined  by 
the  ratio  of  C j to  C^- 


crystal  this  will  be  considerably  greater  than 
the  capacitance  component  of  an  equivalent 
series  L/C  circuit,  and  unless  the  shunt  ca- 
pacitance is  balanced  out  in  a bridge  circuit, 
the  crystal  will  exhibit  both  resonant  (series 
resonant)  and  anti-resonant  (parallel  resonant) 
frequencies,  the  latter  being  slightly  higher 
than  the  series  resonant  frequency  and  ap- 
proaching it  as  Cj  is  increased. 

The  series  resonance  characteristic  is  em- 
ployed in  crystal  filter  circuits  in  receivers 
and  also  in  certain  oscillator  circuits  wherein 
the  crystal  is  used  as  a selective  feedback 
element  in  such  a manner  that  the  phase  of  the 
feedback  is  correct  and  the  amplitude  ade- 
quate only  at  or  very  close  to  the  series  reso- 
nant frequency  of  the  crystal. 

While  quartz,  tourmaline,  Rochelle  salts, 
ADP,  and  EDT  crystals  all  exhibit  the  piezo- 
electric effect,  quartz  is  the  material  widely 
employed  for  frequency  control. 

As  the  cutting  and  grinding  of  quartz  plates 
has  progressed  to  a high  state  of  development 
and  these  plates  may  be  purchased  at  prices 
which  discourage  the  cutting  and  grinding  by 
simple  hand  methods  for  one’s  own  use,  the 
procedure  will  be  only  lightly  touched  upon 
here. 

The  crystal  blank  is  cut  from  the  taw  quartz 
at  a predetermined  orientation  with  respect  to 
the  optical  and  electrical  axes,  the  orientation 
determining  the  activity,  temperature  coeffi- 
cient, thickness  coefficient,  and  other  charac- 
teristics. Various  orientations  or  "cuts”  hav- 
ing useful  characteristics  are  illustrated  in 
figure  5. 


HANDBOOK 


Crystal  Oscillators  245 


Figure  5 

ORIENTATION  OF  THE 
COMMON  CRYSTAL  CUTS 


The  crystal  blank  is  then  rough-ground  al- 
most to  frequency,  the  frequency  increasing 
in  inverse  ratio  to  the  oscillating  dimension 
(usually  the  thickness).  It  is  then  finished  to 
exact  frequency  either  by  careful  lapping,  by 
etching,  or  plating.  The  latter  process  con- 
sists of  finishing  it  to  a frequency  slightly 
higher  than  that  desired  and  then  silver  plating 
the  electrodes  right  on  the  crystal,  the  fre- 
quency decreasing  as  the  deposit  of  silver  is 
increased.  If  the  crystal  is  not  etched,  it  must 
be  carefully  scrubbed  and  "baked”  several 
times  to  stabilize  it,  or  otherwise  the  frequen- 
cy and  activity  of  the  crystal  will  change  with 
time.  Irradiation  by  X-rays  recently  has  been 
used  in  crystal  finishing. 

Unplated  crystals  usually  are  mounted  in 
pressure  holders,  in  which  two  electrodes  are 
held  against  the  crystal  faces  under  slight 
pressure.  Unplated  crystals  also  ate  some- 
times mounted  in  an  air-gap  holder,  in  which 
there  is  a very  small  gap  between  the  crystal 
and  one  or  both  electrodes.  By  making  this 
gap  variable,  the  frequency  of  the  crystal  may 
be  altered  over  narrow  limits  (about  0.3%  for 
certain  types). 

The  temperature  coefficient  of  frequency  for 
various  crystal  cuts  of  the  "-T”  rotated  fam- 
ily is  indicated  in  figure  5.  These  angles  are 
typical,  but  crystals  of  a certain  cut  will  vary 
slightly.  By  controlling  the  orientation  and  di- 
mensioning, the  turning  point  (point  of  zeto 
temperature  coefficient)  for  a BT  cut  plate  may 
be  made  either  lower  or  higher  than  the  75  de- 
grees shown.  Also,  by  careful  control  of  axes 
and  dimensions,  it  is  possible  to  get  AT  cut 

crystals  with  a very  flat  temperature-frequency 
characteristic. 


The  first  quartz  plates  used  were  either  Y 
cut  or  X cut.  The  former  had  a very  high  tem- 
perature coefficient  which  was  discontinuous, 
causing  the  frequency  to  jump  at  certain  criti- 
cal temperatures.  The  X cut  had  a moderately 
bad  coefficient,  but  it  was  more  continuous, 
eind  by  keeping  the  crystal  in  a temperature 
controlled  oven,  a high  order  of  stability  could 
be  obtained.  However,  the  X cut  crystal  was 
considerably  less  active  than  the  Y cut,  es- 
pecially in  the  case  of  poorly  ground  plates. 

For  frequencies  between  500  kc.  and  about 
6 Me.,  the  AT  cut  crystal  now  is  the  most 
widely  used.  It  is  active,  can  be  made  free 
from  spurious  responses,  and  has  an  excellent 
temperature  characteristic.  However,  above 
about  6 Me.  it  becomes  quite  thin,  and  a diffi- 
cult production  job.  Between  6 Me.  and  about 
12  Me.,  the  BT  cut  plate  is  widely  used.  It 
also  works  well  between  500  kc.  and  6 Me., 
but  the  AT  cut  is  mote  desirable  when  a high 
order  of  stability  is  desired  and  no  crystal 
oven  is  employed. 

For  low  frequency  operation  on  the  order  of 
100  kc.,  such  as  is  required  in  a frequency 
standard,  the  GT  cut  crystal  is  recommended, 
though  CT  and  DT  cuts  also  ate  widely  used 
for  applications  between  50  and  500  kc.  The 
CT,  DT,  and  GT  cut  plates  ate  known  as  con- 
tour cuts,  as  these  plates  oscillate  along  the 
long  dimension  of  the  plate  or  bar,  and  are 
much  smaller  physically  than  would  be  the 
case  for  a regular  AT  or  BT  cut  crystal  for 
the  same  frequency. 

Crystal  Holders  Crystals  normally  are  pur- 
chased ready  mounted.  The 


246  Generation  of  R-F  Energy 


THE  RADIO 


best  type  mount  is  determined  by  the  type  crys- 
tal and  its  application,  and  usually  an  opti- 
mum mounting  is  furnished  with  the  crystal. 
However,  certain  features  are  desirable  in  all 
holders.  One  of  these  is  exclusion  of  moisture 
and  prevention  of  electrode  oxidization.  The 
best  means  of  accomplishing  this  is  a metal 
holder,  hermetically  sealed,  with  glass  insula- 
tion and  a metal-to-glass  bond.  However,  such 
holders  are  more  expensive,  and  a ceramic  or 
phenolic  holder  with  rubber  gasket  will  serve 
where  requirements  are  not  too  exacting. 

Temperature  Control;  Where  the  frequency  tol- 

Crystol  Ovens  erance  requirements  are 

not  too  stringent  and  the 
ambient  temperature  does  not  include  extremes, 
an  AT-cut  plate,  or  a BT-cut  plate  with  opti- 
mum (mean  temperature)  turning  point,  will 
often  provide  adequate  stability  without  re- 
sorting to  a temperature  controlled  oven.  How- 
ever, for  broadcast  stations  and  other  applica- 
tions where  very  close  tolerances  must  be 
maintained,  a thermostatically  controlled  oven, 
adjusted  for  a temperature  slightly  higher  than 
the  highest  ambient  likely  to  be  encountered, 
must  of  necessity  be  employed. 

Harmonic  Cut  Just  as  a vibrating  string  can 
Crystals  be  made  to  vibrate  on  its  har- 

monics, a quartz  crystal  will 
exhibit  mechanical  tisonance  (and  therefore 
electrical  resonance)  at  harmonics  of  its  funda- 
mental frequency.  When  employed  in  the  usual 
holder,  it  is  possible  to  excite  the  crystal 
only  on  its  odd  hatmonics  (overtones). 

By  grinding  the  crystal  especially  for  har- 
monic operation,  it  is  possible  to  enhance  its 
operation  as  a harmonic  resonator.  BT  and  AT 
cut  crystals  designed  for  optimum  operation 
on  the  3d,  5 th  and  even  the  7th  harmonic  are 
available.  The  5th  and  7th  harmonic  types, 
especially  the  latter,  require  special  holder 
and  oscillator  circuit  precautions  for  satis- 
factory operation,  but  the  3d  harmonic  type 
needs  little  more  consideration  than  a regular 
fundamental  type.  A crystal  ground  for  optimum 
operation  on  a particular  harmonic  may  or  may 
not  be  a good  oscillator  on  a different  har- 
monic or  on  the  fundamental.  One  interesting 
characteristic  of  a harmonic  cut  crystal  is  that 
its  harmonic  frequency  is  not  quite  an  exact 
multiple  of  its  fundamental,  though  the  dis- 
parity is  very  small. 

The  harmonic  frequency  for  which  the  crys- 
tal was  designed  is  the  working  frequency.  It 
is  not  the  fundamental  since  the  crystal  itself 
actually  oscillates  on  this  working  frequency 
when  it  is  functioning  in  the  proper  manner. 

When  a harmonic-cut  crystal  is  employed,  a 
selective  tuned  circuitmust  be  employed  some- 
where in  the  oscillator  in  order  to  discrimi- 


EXCITATION 


BASIC ‘PIERCE"  OSCILLATOR  HOT-CATHODE  "PIERCE' 
OSCILLATOR 

Figure  6 

THE  PIERCE  CRYSTAL  OSCILLATOR 
CIRCUIT 

Shown  at  (A)  is  the  basic  Pierce  crystal  os- 
cillator circuit.  A capacitance  of  10  to  75 
Pfifd.  normally  will  be  required  at  C j for 
optimum  operation.  If  a plate  supply  voltage 
higher  than  indicated  is  to  be  used,  RFC i 
may  be  replaced  by  a 22,000-ohm  2-watt  re- 
sistor. Shown  at  (B)  is  an  alternative  ar- 
rangement with  the  r-f  ground  moved  to  the 
plate,  and  with  the  cathode  floating.  This 
alternative  circuit  has  the  advantage  that 
the  full  r-f  voltage  developed  across  the 
crystal  may  be  used  as  excitation  to  the  next 
stage,  since  one  side  of  the  crystal  is 
grounded. 


nate  against  the  fundamental  frequency  or  un- 
desired harmonics.  Otherwise  the  crystal  might 
notalways  oscillate  on  the  intended  frequency. 
For  this  reason  the  Pierce  oscillator,  later 
described  in  this  chapter,  is  not  suitable  for 
use  with  harmonic-cut  crystals,  because  the 
only  tuned  element  in  this  oscillator  circuit 
is  the  crystal  itself. 

Crystal  Current;  For  a given  crystal  op- 

Heoting  and  Fracture  erating  as  an  anti-reso- 
nant tank  in  a given  os- 
cillator at  fixed  load  impedance  and  plate  and 
screen  voltages,  the  r-f  current  through  the 
crystal  will  increase  as  the  shunt  capacitance 
Cj  of  figure  4 is  increased,  because  this  effec- 
tively increases  the  step-up  ratio  of  to  Cj. 
For  a given  shunt  capacitance,  Cj,  the  crystal 
current  for  a given  crystal  is  directly  propor- 
tional to  the  r-f  voltage  across  Cj.  This  volt- 
age may  be  measured  by  means  of  a vacuum 
tube  voltmeter  having  a low  input  capacitance, 
and  such  a measurement  is  a more  pertinent 
one  than  a reading  of  r-f  current  by  means  of 
a thermogalvanometer  inserted  in  series  with 
one  of  the  leads  to  the  crystal  holder. 

The  function  of  a crystal  is  to  provide  accu- 
rate frequency  control,  and  unless  it  is  used 
in  such  a manner  as  to  take  advantage  of  its 
inherent  high  stability,  there  is  no  point  in 
using  a crystal  oscillator.  For  this  reason  a 


HANDBOOK 


Crystal  Oscillators  247 


crystal  oscillator  should  not  be  run  at  high 
plate  input  in  an  attempt  to  obtain  consider- 
able power  directly  out  of  the  oscillator,  as 
such  operation  will  cause  the  crystal  to  heat, 
with  resultant  frequency  drift  and  possible 
fracture. 


13-3  Crystal  Oscillator 

Circuits 

Considerable  confusion  exists  as  to  nomen- 
clature of  crystal  oscillator  circuits,  due  to  a 
tendency  to  name  a circuit  after  its  discoverer. 
Nearly  all  the  basic  crystal  oscillator  circuits 
were  either  first  used  or  else  developed  inde- 
pendently by  G.  W.  Pierce,  but  he  has  not  been 
so  credited  in  all  the  literature. 

Use  of  the  crystal  oscillator  in  master  os- 
cillator circuits  in  radio  transmitters  dates 
back  to  about  1924  when  the  first  application 
articles  appeared. 

The  Pierce  The  circuit  of  figure  6A  is  the  sim- 
Oscilletor  plest  crystal  oscillator  circuit.  It 
is  one  of  those  developed  by 
Pierce,  and  is  generally  known  among  ama- 
teurs as  the  Pierce  oscillator.  The  crystal 
simply  replaces  the  tank  circuit  in  a Colpitts 
or  ultra-audion  oscillator.  The  r-f  excitation 
voltage  available  to  the  next  stage  is  low,  be- 
ing somewhat  less  than  that  developed  across 
the  crystal.  Capacitor  C,  will  make  more  of 
the  voltage  across  the  crystal  available  for 
excitation,  and  sometimes  will  be  found  nec- 
essary to  ensure  oscillation.  Its  value  is  small, 
usually  approximately  equal  to  or  slightly 
greater  than  the  stray  capacitance  from  the 
plate  circuit  to  ground  (including  the  grid  of 
the  stage  being  driven). 

If  the  r-f  choke  has  adequate  inductance,  a 
crystal  (even  a harmonic  cut  crystal)  will  al- 
most invariably  oscillate  on  its  fundamental. 
The  Pierce  oscillator  therefore  cannot  be  used 
with  harmonic  cut  crystals. 

The  circuit  at  (B)  is  the  same  as  that  of 
(A)  except  that  the  plate  instead  of  the  cath- 
ode is  operated  at  ground  r-f  potential.  All  of 
the  r-f  voltage  developed  across  the  crystal 
is  available  for  excitation  to  the  next  stage, 
but  still  is  low  for  reasonable  values  of  crys- 
tal current.  For  best  operation  a tube  with  low 
heater-cathode  capacitance  is  required.  Exci- 
tation for  the  next  stage  may  also  be  taken 
from  the  cathode  when  using  this  circuit. 

Tuned-Plate  The  circuit  shown  in  fig- 

Crystal  Oscillator  ure  7A  is  also  one  used  by 
Pierce,  but  is  more  widely 
referred  to  as  the  "Miller”  oscillator.  To  avoid 


confusion,  we  shall  refer  to  it  as  the  tuned- 
plate  crystal  oscillator.  It  is  essentially  an 
Armstrong  or  tuned  plate-tuned  grid  oscillator 
with  the  crystal  replacing  the  usual  L-C  grid 
tank.  The  plate  tank  must  be  tuned  to  a fre- 
quency slightly  higher  than  the  anti-resonant 
(parallel  resonant)  frequency  of  the  crystal. 
Whereas  the  Pierce  circuits  of  figure  6 will 
oscillate  at  (or  very  close  to)  the  anti-reso- 
nant frequency  of  the  crystal,  the  circuits  of 
figure  7 will  oscillate  at  a frequency  a little 
above  the  anti-resonant  frequency  of  rhe 
crystal. 

The  diagram  shown  in  figure  7A  is  the  basic 
circuit.  The  most  popular  version  of  the  tuned- 
plate  oscillator  employs  a pentode  or  beam 
tetrode  with  cathode  bias  to  prevent  excessive 
plate  dissipation  when  the  circuit  is  not  oscil- 
lating. The  cathode  resistor  is  optional.  Its 
omission  will  reduce  both  crystal  current  and 
oscillator  efficiency,  resulting  in  somewhat 
mote  output  for  a given  crystal  current.  The 
tube  usually  is  an  audio  or  video  beam  pentode 
or  tetrode,  the  plate-grid  capacitance  of  such 
tubes  being  sufficient  to  ensure  stable  oscilla- 
tion but  not  so  high  as  to  offer  excessive  feed- 
back with  resulting  high  crystal  current.  The 
6AG7  makes  an  excellent  ^1-around  tube  for 
this  type  citcuit. 

Pentode  The  usual  type  of  Crystal- 

Harmonic  Crystal  controlled  h-f  transmitter 

Oscillator  Circuits  operates,  at  least  part  of 
the  time,  on  a frequency 
which  is  an  integral  multiple  of  the  operating 
frequency  of  the  controlling  crystal.  Hence, 
oscillator  circuits  which  are  capable  of  pro- 
viding output  on  the  ctystal  frequency  if  de- 
sired, but  which  also  can  deliver  output  energy 
on  harmonics  of  the  ctystal  frequency  have 
come  into  wide  use.  Four  such  circuits  which 
have  found  wide  application  are  illustrated  in 
figures  7C,  7D,  7E,  and  7F. 

The  circuit  shown  in  figure  7C  is  recom- 
mended for  use  with  harmonic-cut  crystals 
when  output  is  desired  on  a multiple  of  the 
oscillating  frequency  of  the  crystal.  As  an 
example,  a 25-Mc.  harmonic-cut  ctystal  may 
be  used  in  this  circuit  to  obtain  output  on  50 
Me.,  or  a 48-Mc.  harmonic-cut  crystal  may  be 
used  to  obtain  output  on  the  144-Mc.  amateur 
band.  The  citcuit  is  not  recommended  for  use 
with  the  normal  type  of  fundamental-frequency 
crystal  since  more  output  with  fewer  variable 
elements  can  be  obtained  with  the  circuits  of 
7D  and  7F. 

The  Pietce-hatmonic  circuit  shown  in  fig- 
ure 7D  is  satisfactory  for  many  applications 
which  require  very  low  crystal  current,  but 

has  the  disadvantage  that  both  sides  of  the 
crystal  are  above  ground  potential.  The  Tri-tet 
circuit  of  figure  7E  is  widely  used  and  can 


248  Generation  of  R-F  Energy 


THE  RADIO 


6J5,  ETC. 


IL 


^ .002 


O 

Jy  o 

n o 
o 


BASIC  TUNED-PLATE  OSCILLATOR 

@ 


RECOMMENDED  TUNED-PLATE 
OSCILLATOR 

© 


SPECIAL  CIRCUIT  FOR  USE  WITH 
HARMONIC-CUT  CRYSTAL. 

© 


6AG7  F.  2F,  3F 


© 


6AG7  F.2F,  3F.  4F 


© 


6AG7  F.  2F.  3F.  4F 


© 


Figure  7 

COMMONLY  USED  CRYSTAL  OSCILLATOR  CIRCUITS 

Shown  at  (A)  Is  tho  basic  tuned-plata  crystal  oscillator  with  a triode  oscillator  tube. 
The  plate  tank  must  be  tuned  on  the  low~c<^aeltonce  side  pf  resonance  to  sustain 
oscillation.  (B)  shows  the  tuned-plate  oscillator  as  it  is  normally  used,  with  an  a-i 
power  pentode  to  permit  high  output  with  relatively  Ipw  crystal  current. 
Schematics  (C),  (D),  (E),  and  (F)  Illustrate  crystal  oscillator  circuits  which  can  de- 
liver moderate  output  energy  on  harmonics  of  the  oscillating  frequency  of  the  crys- 
tal. (C)  shows  a special  circuit  which  will  permit  use  of  a harmonic-cut  crystal  to 
obtain  output  energy  well  into  the  v-h-f  range.  (D)  is  valuable  when  extremely  low 
crystal  current  is  a requirement,  but  delivers  relatively  low  output,  (E)  is  commonly 
used,  but  is  subject  to  crystal  damage  if  the  cathode  circuit  is  mistuned.  (F)  is 
recommended  as  the  most  generally  satisfactory  from  the  standpoints  of:  low  crys- 
tal current  regardless  of  mis-adjustment,  good  output  on  harmonic  frequencies,  one 
side  of  crystal  is  grounded,  will  oscillate  with  crystals  from  1,5  to  10  Me.  without 
adjustment,  output  tank  may  be  tuned  to  the  crystal  frequency  for  fundamental  out- 
put without  stopping  oscillation  or  changing  frequency. 


give  excellent  output  with  low  crystal  current. 
However,  the  circuit  has  the  disadvantages  of 
requiring  a cathode  coil,  of  requiring  careful 
setting  of  the  variable  cathode  capacitor  to 
avoid  damaging  the  crystal  when  changing  fre- 
quency ranges,  and  of  having  both  sides  of 
the  crystal  above  ground  potential. 

The  Colpitts  harmonic  oscillator  of  figure 
7F  is  recommended  as  being  the  most  gener- 
ally satisfactory  harmonic  crystal  oscillator 
circuit  since  it  has  the  following  advantages: 
(1)  the  circuit  will  oscillate  with  crystals 
over  a very  wide  frequency  range  with  no 
change  other  than  plugging  in  or  switching  in 
the  desired  crystal;  (2)  crystal  current  is  ex- 


tremely low;  (3)  one  side  of  the  crystal  is 
grounded,  which  facilitates  crystal-switching 
circuits;  (4)  the  circuit  will  operate  straight 
through  without  frequency  pulling,  or  it  may 
be  operated  with  output  on  the  second,  third, 
or  fourth  harmonic  of  the  crystal  frequency. 

Crystal  Oscillator  The  tunable  circuits  of  all 

Tuning  oscillators  illustrated 

should  be  tuned  for  maxi- 
mum output  as  indicated  by  maximum  excita- 
tion to  the  following  stage,  except  that  the 
oscillator  tank  of  tuned-plate  oscillators  (fig- 
ure 7A  and  figure  7B)  should  be  backed  off 
slightly  towards  the  low  capacitance  side  from 


HANDBOOK 


All-band  Crystal  Oscillator  24  9 


MOT-ES 

1.  \-\=l5VH  {2^"  OF  BAW  It  3015) 

2.  V-Z-FSJJH  {!"  OF  BAWHSOOA) 

3.  FOB  no  METEB  OPERATION  ADD  S JJSJF.  CONDENSER 
BETWEEN  PINS  A A B OF  BAST.  PLATE  COIL  = SB  JJN, 

(2  ^"02  BAWttSOIB\ 

4.  X=  7 MC.  CRYSTAL  FOR  HARMONIC  OPERATION 

Figure  8 

ALL-BAND  6AG7  CRYSTAL  OSCILLATOR 
CAPABLE  OF  DRIVING 
BEAM  PENTODE  TUBE 


maximum  output,  as  the  oscillator  then  is  in 
a more  stable  condition  and  sure  to  start  im- 
mediately when  power  is  applied.  This  is  es- 
pecially important  when  the  oscillator  is  keyed, 
as  for  break-in  c-w  operation. 

Crystal  Switching  It  is  desirable  to  keep  stray 
shunt  capacitances  in  the 
crystal  circuit  as  low  as  possible,  regardless 
of  the  oscillator  circuit.  If  a selector  switch 
is  used,  this  means  that  both  switch  and  crys- 
tal sockets  must  be  placed  close  to  the  oscil- 
lator tube  socket.  This  is  especially  true  of 
harmonic-cut  crystals  operating  on  a compara- 
tively high  frequency.  In  fact,  on  the  highest 
frequency  crystals  it  is  preferable  to  use  a 
turret  arrangement  for  switching,  as  the  stray 
capacitances  can  be  kept  lower. 

Crystol  Oscillator  When  the  crystal  oscillator 
Keying  is  keyed,  it  is  necessary 

that  crystal  activity  and  os- 
cillator-tube transconductance  be  moderately 
high,  and  that  oscillator  loading  and  crystal 
shunt  capacitance  be  low.  Below  2500  kc.  and 
above  6 Me.  these  considerations  become  es- 
pecially important.  Keying  of  the  plate  voltage 
(in  the  negative  lead)  of  a crystal  oscillator, 
with  the  screen  voltage  regulated  at  about 
150  volts,  has  been  found  to  give  satisfactory 
results. 

A Versatile  6AG7  The  6AG7  tube  may  be 

Crystal  Oscillator  used  in  a modified  Tti-cec 

crystal  oscillator,  capable 
of  delivering  sufficient  power  on  all  bands 


from  160  meters  through  10  meters  to  fully 
drive  a pentode  tube,  such  as  the  807,  2E26 
or  6146.  Such  an  oscillator  is  extremely  use- 
ful for  portable  or  mobile  work,  since  it  com- 
bines all  essential  exciter  functions  in  one 
tube.  The  circuit  of  this  oscillator  is  shown 
in  figure  8.  For  160,  80  and  40  meter  opera- 
tion the  6AG7  functions  as  a tuned-plate  os- 
cillator. Fundamental  frequency  crystals  are 
used  on  these  three  bands.  For  20,  15  and  10 
meter  operation  the  6AG7  functions  as  a Tri- 
tet  oscillator  with  a fixed-tuned  cathode  cir- 
cuit. The  impedance  of  this  cathode  circuit 
does  not  affect  operation  of  the  6AG7  on  the 
lower  frequency  bands  so  it  is  left  in  the  cir- 
cuit at  all  times.  A 7-Mc.  crystal  is  used  for 
fundamental  output  on  40  meters,  and  for  har- 
monic output  on  20,  15  and  10  meters.  Crystal 
current  is  extremely  low  regardless  of  the  out- 
put frequency  of  the  oscillator.  The  plate  cir- 
cuit of  the  6AG7  is  capable  of  tuning  a fre- 
quency range  of  2:1,  requiring  only  two  output 
coils:  one  for  80-40  meter  operation,  and  one 
for  20,  15  and  10  meter  operation.  In  some 

cases  it  may  be  necessary  to  add  5 micromicro- 
fatads  of  external  feedback  capacity  between 
the  plate  and  control  grid  of  the  6AG7  tube  to 
sustain  oscillation  with  sluggish  160  meter 
crystals. 

Triode  Overtone  The  recent  development  of 

Oscillators  reliable  overtone  crystals 

capable  of  operation  on  the 
third,  fifth,  seventh  (or  higher)  overtones  has 
made  possible  v-h-f  output  from  a low  frequen- 
cy crystal  by  the  use  of  a double  triode  regen- 
erative oscillator  circuit.  Some  of  the  new 
twin  triode  tubes  such  as  the  12AU7,  12AV7 
and  6j6  are  especially  satisfactory  when  used 
in  this  type  of  circuit.  Crystals  that  are  ground 
for  overtone  service  may  be  made  to  oscillate 
on  odd  overtone  frequencies  other  than  the 
one  marked  on  the  crystal  holder.  A 24-Mc. 
overtone  crystal,  for  example,  is  a specially 
ground  8-Mc.  crystal  operating  on  its  third 
overtone.  In  the  proper  circuit  it  may  be  made 
to  oscillate  on  40  Me.  (fifth  overtone),  56  Me. 
(seventh  overtone),  or  72  Me.  (ninth  overtone). 
Even  the  ordinary  8-Mc.  crystals  not  designed 
for  overtone  operation  may  be  made  to  oscil- 
late readily  on  24  Me.  (third  overtone)  in  these 
circuits. 

A variety  of  overtone  oscillator  circuits  is 
shown  in  figure  9.  The  oscillator  of  figure  9A 
is  attributed  to  Frank  Jones,  W6AJF.  The  first 
section  of  the  6j6  dual  triode  comprises  a re- 
generative oscillator,  with  output  on  either  the 
third  or  fifth  overtone  of  the  crystal  frequency. 
The  regenerative  loop  of  this  oscillator  con- 
sists of  a condenser  bridge  made  up  of  and 
Cj,  with  the  ratio  Q.J C,  determining  the  amount 
of  regenerative  feedback  in  the  circuit.  With 


250  Generation  of  R-F  Energy 


THE  RADIO 


6J6 

3FOR5F 


@ JONES  HARMONIC  OSCILLATOR 


@ COLPITTS  HARMONIC  OSCILLATOR 


6,  9F 


THSSe  COILS  MAD£  FROM  SINGLE  SECTION 
OF  MINIDUCTOfK  ONE  TURN  BROKEN  TO 
OlVIDS  INDUCTOR  INTO  TWO  COILS. 


(C)  REGENERATIVE  HARMONIC  OSCILLATOR 


6J6 


+300 V Li  = /or  # son  bsw  m/n/ductor. 

TAPAT3T.  FROM  GRID  END 


@ REGENERATIVE  HARMONIC  OSCILLATOR 


■^laATT  (or  6AB4) 


+ 200  V. 


Li-Sr.lt/$,  D.  GRACED 
L2S/r  HOOKUR  WIRE,  D. 


© CATHODE  FOLLOWER  OVERTONE  OSCILLATOR 


© V.H.F.  OVERTONE  OSCILLATOR 


Figure  9 

VARIOUS  TYPES  OF  OVERTONE  OSCILLATORS  USING  MINIATURE  DOUBL E-TRIODE 

TUBES 


an  8-Mc.  crystal,  output  from  the  first  section 
of  the  6j6  tube  may  be  obtained  on  either  24 
Me.  or  40  Me.,  depending  upon  the  resonant 
frequency  of  the  plate  circuit  inductor,  L,.  The 
second  half  of  the  6j6  acts  as  a frequency 
multiplier,  its  plate  circuit,  Lj,  tuned  to  the 
sixth  or  ninth  harmonic  frequency  when  L,  is 
tuned  to  the  third  overtone,  or  to  the  tenth 
harmonic  frequency  when  Lj  is  tuned  to  the 
fifth  overtone. 

Figure  9B  illustrates  a Colpitts  overtone 
oscillator  employing  a 6j6  tube.  This  is  an 
outgrowth  of  the  Colpitts  harmonic  oscillator 
of  figure  7F.  The  regenerative  loop  in  this 


case  consists  of  C,,  Cj  and  RFC  between  the 
grid,  cathode  and  ground  of  the  first  section 
of  the  6j6.  The  plate  circuit  of  the  first  sec- 
tion is  tuned  to  the  second  overtone  of  the 
crystal,  and  the  second  section  of  the  6j6 
doubles  to  the  fourth  harmonic  of  the  crystal. 
This  circuit  is  useful  in  obtaining  28-Mc.  out- 
put from  a 7-Mc.  crystal  and  is  highly  popular 
in  mobile  work. 

The  circuit  of  figure  9C  shows  a typical  re- 
generative overtone  oscillator  employing  a 
12AU7  double  triode  tube.  Feedback  is  con- 
trolled by  the  number  of  turns  in  Lj,  and  the 
coupling  between  L,  and  L,.Only  enough  feed- 


HANDBOOK 


R-F  Amplifiers  251 


back  should  be  employed  to  maintain  proper 
oscillation  of  the  crystal.  Excessive  feedback 
will  cause  the  first  section  of  the  12AU7  to 
oscillate  as  a self-excited  TNT  oscillator, 
independent  of  the  crystal.  A variety  of  this 
circuit  is  shown  in  figure  9D,  wherein  a tapped 
coil,  Li,  is  used  in  place  of  the  two  separate 
coils.  Operation  of  the  circuit  is  the  same  in 
either  case,  regeneration  now  being  controlled 
by  the  placement  of  the  tap  on  L^. 

A cathode  follower  overtone  oscillator  is 
shown  in  figure  9E.  The  cathode  coil,  L,,  is 
chosen  so  as  to  resonate  with  the  crystal  and 
tube  capacities  just  below  the  third  overtone 
frequency  of  the  crystal.  For  example,  with  an 
8-Mc.  crystal,  Lj  is  tuned  to  24  Me.,  Li  reson- 
ates with  the  circuit  capacities  to  23*5  Me., 
and  the  harmonic  tank  circuit  of  the  second 
section  of  the  12AT7  is  tuned  either  to  48  Me. 
or  72  Me.  If  a 24-Mc.  overtone  crystal  is  used 
in  this  circuit,  La  may  be  tuned  to  72  Me.,  Lj 
resonates  with  the  circuit  capacities  to  70 
Me.,  and  the  harmonic  tank  circuit,  Lj,  is  tuned 
to  144  Me.  If  there  is  any  tendency  towards 
self-oscillation  in  the  circuit,  it  may  be  elimi- 
nated by  a small  amount  of  inductive  coupling 
between  L,  and  L3.  Placing  these  coils  near 
each  other,  with  the  winding  of  Lj  correctly 
polarized  with  respect  to  Lj  will  prevent  self- 
oscillation of  the  circuit. 

The  use  of  a 144-Mc.  overtone  crystal  is 
illustrated  in  figure  9F,  A 6AB4  or  one-half 
of  a 12AT7  tube  may  be  used,  with  output 
directly  in  the  2-meter  amateur  band.  A slight 
amount  of  regeneration  is  provided  by  the  one 
turn  link,  Lj,  which  is  loosely  coupled  to  the 
144-Mc.  tuned  tank  circuit,  Li  in  the  plate  cir- 
cuit of  the  oscillator  tube.  If  a 12AT7  tube 
and  a 110-Mc.  crystal  are  employed,  direct  out- 
put in  the  220-Mc.  amateur  band  may  be  ob- 
tained from  the  second  half  of  the  12AT7. 


13-4  Radio  Frequency 

Amplifiers 

The  output  of  the  oscillator  stage  in  a trans- 
mitter (whether  it  be  self-controlled  or  crystal 
controlled)  must  be  kept  down  to  a fairly  low 
level  to  maintain  stability  and  to  maintain  a 
factor  of  safety  from  fracture  of  the  crystal 
when  one  is  used.  The  low  power  output  of 
the  oscillator  is  brought  up  to  the  desired 
power  level  by  means  of  radio-frequency  am- 
plifiers. The  two  classes  of  r-f  amplifiers  that 
find  widest  application  in  radio  transmitters 
are  the  Class  B and  Class  C types. 

The  Class  B Class  B amplifiers  are  used  in  a 

Amplifier  radio-telegraph  transmitter  when 
maximum  power  gain  and  mini- 


mum harmonic  output  is  desired  in  a particu- 
lar stage.  A Class  B amplifier  operates  with 
cutoff  bias  and  a comparatively  small  amount 
of  excitation.  Power  gains  of  20  to  200  or  so 
ate  obtainable  in  a well-designed  Class  B 
amplifier.  The  plate  efficiency  of  a Class  B 
c-w  amplifier  will  tun  around  65  per  cent. 

The  Class  B Another  type  of  Class  B ampli- 
Linear  fiet  is  the  Class  B linear  stage 

as  employed  in  radiophone  work. 
This  type  of  amplifier  is  used  to  increase  the 
level  of  a modulated  carrier  wave,  and  de- 
pends for  its  operation  upon  the  linear  rela- 
tion between  excitation  voltage  and  output 
voltage.  Or,  to  state  the  fact  in  another  man- 
ner, the  power  output  of  a Class  B linear  stage 
varies  linearly  with  the  square  of  the  excita- 
tion voltage. 

The  Class  B linear  amplifier  is  operated 
with  cutoff  bias  and  a small  value  of  excita- 
tion, the  actual  value  of  exciting  power  being 
such  that  the  power  output  under  carrier  con- 
ditions is  one-fourth  of  the  peak  power  capa- 
bilities of  the  stage.  Class  B linears  are  very 
widely  employed  in  broadcast  and  commercial 
installations,  but  are  comparatively  uncommon 
in  amateur  application,  since  tubes  with  high 
plate  dissipation  are  required  for  moderate 
output.  The  carrier  efficiency  of  such  an  am- 
plifier will  vary  from  approximately  30  per 
cent  to  35  per  cent. 

The  Class  C Class  C amplifiers  are  very  wide- 
^ Amplifier  ly  used  in  all  types  of  trans- 
mitters. Good  power  gain  may  be 
obtained  (values  of  gain  from  3 to  20  are  com- 
mon) and  the  plate  circuit  efficiency  may  be, 
under  certain  conditions,  as  high  as  85  per 
cent.  Class  C amplifiers  operate  with  consider- 
ably more  than  cutoff  bias  and  ordinarily  with 
a large  amount  of  excitation  as  compared  to  a 
Class  B amplifier.  The  bias  for  a normal  Class 
C amplifier  is  such  that  plate  current  on  the 
stage  flows  for  approximately  120°  of  the  360° 
excitation  cycle.  Class  C amplifiers  are  used 
in  transmitters  where  a fairly  large  amount  of 
excitation  power  is  available  and  good  plate 
circuit  efficiency  is  desired. 

Plate  Modulated  The  characteristic  of  a Class 
Class  C C amplifier  which  makes  it 

linear  with  respect  to 
changes  in  plate  voltage  is  that  which  allows 
such  an  amplifier  to  be  plate  modulated  for 
radiotelephony.  Through  the  use  of  higher  bias 
than  is  required  for  a c-w  Class  C amplifier 
and  greater  excitation,  the  linearity  of  such 
an  amplifier  may  be  extended  from  zero  plate 
voltage  to  twice  the  normal  value.  The  output 
power  of  a Class  C amplifier,  adjusted  for 
plate  modulation,  varies  with  the  square  of  the 


252  Generation  of  R-F  Energy 


THE  RADIO 


plate  voltage.  This  is  the  same  condition  that 
would  take  place  if  a resistor  equal  to  the 
voltage  on  the  amplifier,  divided  by  its  plate 
current,  were  substituted  for  the  amplifier. 
Therefore,  the  stage  presents  a resistive  load 
to  the  modulator. 

Grid  Modulated  If  the  grid  current  to  a Class 
Class  C C amplifier  is  reduced  to  a 

low  value,  and  the  plate  load- 
ing is  increased  to  the  point  where  the  plate 
dissipation  approaches  the  rated  value,  such 
an  amplifier  may  be  grid  modulated  for  radio- 
telephony. If  the  plate  voltage  is  raised  to 
quite  a high  value  and  the  stage  is  adjusted 
carefully,  efficiencies  as  high  as  40  to  43  per 
cent  with  good  modulation  capability  and  com- 
paratively low  distortion  may  be  obtained. 
Fixed  bias  is  required.  This  type  of  operation 
is  termed  Class  C grid-bias  modulation. 

Grid  Excitatlan  Adequate  grid  excitation 
must  be  available  for  Class 
B or  Class  C service.  The  excitation  for  a 
plate-modulated  Class  C stage  must  be  suffi- 
cient to  produce  a normal  value  of  d-c  grid  cur- 
rent with  rated  bias  voltage.  The  bias  voltage 
preferably  should  be  obtained  from  a combina- 
tion of  grid  leak  and  fixed  C-bias  supply. 

Cutoff  bias  can  be  calculated  by  dividing 
the  amplification  factor  of  the  tube  into  the 
d-c  plate  voltage.  This  is  the  value  normally 
used  for  Class  B amplifiers  (fixed  bias,  no 
grid  resistor).  Class  C amplifiers  use  from  114 
to  5 times  this  value,  depending  upon  the  avail- 
able grid  drive,  or  excitation,  and  the  desired 
plate  efficiency.  Less  grid  excitation  is  need- 
ed for  c-w  operation,  and  the  values  of  fixed 
bias  (if  greater  than  cutoff)  may  be  reduced,  or 
the  value  of  the  grid  leak  resistor  can  be  low- 
ered until  normal  rated  d-c  grid  current  flows. 

The  values  of  grid  excitation  listed  for  each 
type  of  tube  may  be  reduced  by  as  much  as 
50  per  cent  if  only  moderate  power  output  and 
plate  efficiency  are  desired.  When  consulting 
the  tube  tables,  it  is  well  to  remember  that 
the  power  lost  in  the  tuned  circuits  must  be 
taken  into  consideration  when  calculating  the 
available  grid  drive.  At  very  high  frequencies, 
the  r-f  circuit  losses  may  even  exceed  the 
power  required  for  actual  grid  excitation. 

Link  coupling  between  stages,  particularly 
to  the  final  amplifier  grid  circuit,  normally  will 
provide  more  grid  drive  than  can  be  obtained 
from  other  coupling  systems.  The  number  of 
turns  in  the  coupling  link,  and  the  location  of 
the  turns  on  the  coil,  can  be  varied  with  res- 
pect to  the  tuned  circuits  to  obtain  the  great- 
est grid  drive  for  allowable  values  of  buffer 
or  doubler  plate  current.  Slight  readjustments 
sometimes  can  be  made  after  plate  voltage 
has  been  applied  to  the  driver  tube. 


Excessive  grid  current  damages  tubes  by 

overheating  the  grid  structure;  beyond  a cer- 
tain point  of  grid  drive,  no  increase  in  power 
output  can  be  obtained  for  a given  plate  volt- 
age. 

13-5  Neutralization  of 

R.F.  Amplifiers 

The  plate-to-grid  feedback  capacitance  of 
triodes  makes  it  necessary  that  they  be  neu- 
tralized for  operation  as  r-f  amplifiers  at  fre- 
quencies above  about  500  kc.  Those  screen- 
grid  tubes,  pentodes,  and  beam  tetrodes  which 
have  a plate-to-grid  capacitance  of  0.1 
or  less  may  be  operated  as  an  amplifier  with- 
out neutralization  in  a well-designed  amplifier 
up  to  30  Me. 

Neutralizing  The  object  qf  neutralization  is 
Circuits  to  cancel  or  neutralize  the  ca- 

pacitive feedback  of  energy  from 
plate  to  grid.  There  are  two  general  methods 
by  which  this  energy  feedback  may  be  elimi- 
nated: the  first,  and  the  most  common  method, 
is  through  the  use  of  a capacitance  bridge, 
and  the  second  method  is  through  the  use  of  a 
parallel  reactance  of  equal  and  opposite  po- 
larity to  the  grid-to-plate  capacitance,  to  nul- 
lify the  effect  of  this  capacitance. 

Examples  of  the  first  method  are  shown  in 
figure  10.  Figure  lOA  shows  a capacity  neu- 
tralized stage  employing  a balanced  tank  cir- 
cuit. Phase  reversal  in  the  tank  circuit  is  ob- 
tained by  grounding  the  center  of  the  tank  coil 
to  radio  frequency  energy  by  condenser  C. 
Points  A and  B are  180  degrees  out  of  phase 
with  each  other,  and  the  correct  amount  of  out 
of  phase  energy  is  coupled  through  the  neu- 
tralizing condenser  NC  to  the  grid  circuit  of 
the  tube.  The  equivalent  bridge  circuit  of  this 
is  shown  in  figure  llA.  It  is  seen  that  the 
bridge  is  not  in  balance,  since  the  plate-fila- 
ment capacity  of  the  tube  forms  one  leg  of  the 
bridge,  and  there  is  no  corresponding  capacity 
from  the  neutralizing  condenser  (point  B)  to 
ground  to  obtain  a complete  balance.  In  addi- 
tion, it  is  mechanically  difficult  to  obtain  a 
perfect  electrical  balance  in  the  tank  coil,  and 
the  potential  between  point  A and  ground  and 
point  B and  ground  in  most  cases  is  unequal. 
This  circuit,  therefore,  holds  neutralization 
over  a very  small  operating  range  and  unless 
tubes  of  low  interelectrode  capacity  are  used 
the  inherent  unbalance  of  the  circuit  will  per- 
mit only  approximate  neutralization. 

Split-Statar  Figure  lOB  shows  the  neu- 

Plate  Neutral  I-  tralization  circuit  which  is 
zation  most  widely  used  in  single- 

ended  r-f  stages.  The  use  of 


HANDBOOK 


Neutral  i zation  253 


A 

IT 


^1^ 


© 


© 


Figure  10 

COMMON  NEUTRALIZING  CIRCUITS  FOR  SINGLE-ENDED  AMPLIFIERS 


a split-stator  plate  capacitor  makes  the  electri- 
cal balance  of  the  circuit  substantially  inde- 
pendent of  the  mutual  coupling  within  the  coil 
and  also  makes  the  balance  independent  of  the 
place  where  the  coil  is  tapped.  With  conven- 
tional tubes  this  circuit  will  allow  one  neutral- 
ization adjustment  to  be  made  on,  say,  28  Me., 
and  this  adjustment  usually  will  hold  suffi- 
ciently close  for  operation  on  all  lower  fre- 
quency bands. 

Condenser  C,  is  used  to  balance  out  the 
plate-filament  capacity  of  the  tube  to  allow  a 
perfect  neutralizing  balance  at  all  frequencies. 
The  equivalent  bridge  circuit  is  shown  in  fig- 
ure IIB.  If  the  plate-filament  capacity  of  the 
tube  is  extremely  low  (lOOTH  triode,  for  ex- 
ample), condenser  C,  may  be  omitted,  or  may 
merely  consist  of  the  residual  capacity  of  NC 
to  ground. 

Grid  Neutralization  A split  grid  tank  circuit 
may  also  be  used  for  neu- 
tralization of  a triode  tube  as  shown  in  figure 
IOC.  Out  of  phase  voltage  is  developed  across 
a balanced  grid  circuit,  and  coupled  through 
NC  to  the  single-ended  plate  circuit  of  the 
tube.  The  equivalent  bridge  circuit  is  shown 
in  figure  IIC.  This  circuit  is  in  balance  until 
the  stage  is  in  operation  when  the  loading  ef- 
fect of  the  tube  upon  one-half  of  the  grid  cir- 
cuit throws  the  bridge  circuit  out  of  balance. 

The  amount  of  unbalance  depends  upon  the 

grid-plate  capacity  of  the  tube,  and  the  amount 
of  mutual  inductance  between  the  two  halves 


of  the  grid  coil.  If  an  t-f  voltmeter  is  placed 
between  point  A and  ground,  and  a second 
voltmeter  placed  between  point  B and  ground 
the  loading  effect  of  the  tube  will  be  notice- 
able. When  the  tube  is  supplied  excitation 
with  no  plate  voltage,  NC  may  be  adjusted 
until  the  circuit  is  in  balance.  When  plate 
voltage  is  applied  to  the  stage,  the  voltage 
from  point  A to  ground  will  decrease,  and  the 
voltage  from  point  B to  ground  will  increase, 
both  in  direct  proportion  to  the  amount  of  cir- 
cuit unbalance.  The  use  of  this  circuit  is  not 
recommended  above  7 Me.,  and  it  should  be 
used  below  that  frequency  only  with  low  in- 
ternal capacity  tubes. 

Push-Pull  Two  tubes  of  the  same  type 

Neutralization  can  be  connected  for  push-pull 
operation  so  as  to  obtain  twice 
as  much  output  as  that  of  a single  tube.  A 
push-pull  amplifier,  such  as  that  shown  in  fig- 
ure 12  also  has  an  advantage  in  that  the  cir- 
cuit can  mote  easily  be  balanced  than  a single- 
tube r-f  amplifier.  The  various  inter-electrode 
capacitances  and  the  neutralizing  capacitors 
are  connected  in  such  a manner  that  the  re- 
actances on  one  side  of  the  tuned  circuits  ate 
exactly  equal  to  those  on  the  opposite  side. 
For  this  reason,  push-pull  r-f  amplifiers  can 
be  more  easily  neutralized  in  very-high-fre- 
quency transmitters;  also,  they  usually  remain 

in  perfect  neutralization  when  tuning  the  am- 

plifier to  different  bands. 

The  circuit  shown  in  figure  12  is  perhaps 


254  Generation  of  R-F  Energy 


THE  RADIO 


c 


@ BRIDGE  EQUIVALENT  OF  FIGURE  lO-A 


c 


@ BRIDGE  EQUIVALENT  OF  FIGURE  lO-B 


C 


© BRIDGE  EQUIVALENT  OF  FIGURE  lO-C 


Figure  1 1 

EQUIVALENT  NEUTRALIZING  CIRCUITS 


the  most  commonly  used  arrangement  for  a 
push-pull  r-f  amplifier  stage.  The  rotor  of  the 
grid  capacitor  is  grounded,  and  the  rotor  of  the 
plate  tank  capacitor  is  by-passed  to  ground. 

Shunt  or  Coil  The  feedback  of  energy  from 
Neutralization  grid  to  plate  in  an  unneutral- 
ized r-f  amplifier  is  a result  of 
the  grid-to-plate  capacitance  of  the  amplifier 
tube.  A neutralization  circuit  is  merely  an 
electrical  arrangement  for  nullifying  the  effect 
of  this  capacitance.  All  the  previous  neutrali- 
zation circuits  have  made  use  of  a bridge  cir- 
cuit for  balancing  out  the  grid-to-plate  energy 
feedback  by  feeding  back  an  equal  amount  of 
energy  of  opposite  phase. 

Another  method  of  eliminating  the  feedback 
effect  of  this  capacitance,  and  hence  of  neu- 
tralizing the  amplifier  stage,  is  shown  in  fig- 
ure 13.  The  grid-to-plate  capacitance  in  the 
triode  amplifier  tube  acts  as  a capacitive  re- 


Figure 12 

STANDARD  CROSS-NEUTRALIZED 
PUSH-PULL  TRIODE  AMPLIFIER 


actance,  coupling  energy  back  from  the  plate 
to  the  grid  circuit.  If  this  capacitance  is  par- 
alleled with  an  inductance  having  the  same 
value  of  reactance  of  opposite  sign,  the  re- 
actance of  one  will  cancel  the  reactance  of 
the  other  and  a high-impedance  tuned  circuit 
from  grid  to  plate  will  result. 

This  neutralization  circuit  can  be  used  on 
ultra-high  frequencies  where  other  neutraliza- 
tion circuits  are  unsatisfactory.  This  is  true 
because  the  lead  length  in  the  neutralization 
circuit  is  practically  negligible.  The  circuit 
can  also  be  used  with  push-pull  r-f  amplifiers. 
In  this  case,  each  tube  will  have  its  own  neu- 
tralizing inductor  connected  from  grid  to  plate. 

The  main  advantage  of  this  arrangement  is 
that  it  allows  the  use  of  single-ended  tank 
circuits  with  a single-ended  amplifier. 

The  chief  disadvantage  of  the  shunt  neutral- 
ized arrangement  is  that  the  stage  must  be  re- 
neutralized each  time  the  stage  is  retuned  to 
a new  frequency  sufficiently  removed  that  the 
grid  and  plate  tank  circuits  must  be  retuned  to 
resonance.  However,  by  the  use  of  plug-in 
coils  it  is  possible  to  change  to  a different 
band  of  operation  by  changing  the  neutral- 
izing coil  at  the  same  time  that  the  grid  and 
plate  coils  are  changed. 

The  0,0001-fild.  capacitor  in  series  with 
the  neutralizing  coil  is  merely  a blocking  ca- 
pacitor to  isolate  the  plate  voltage  from  the 
grid  circuit.  The  coil  L will  have  to  have  a 
very  large  number  of  turns  for  the  band  of  oper- 
ation in  order  to  be  resonant  with  the  compara- 
tively small  grid-to-plate  capacitance.  But 
since,  in  all  ordinary  cases  with  tubes  operat- 
ing on  frequencies  for  which  they  were  de- 
signed, the  L/C  ratio  of  the  tuned  circuit  will 
be  very  high,  the  coil  can  use  comparatively 
small  wire,  although  it  must  be  wound  on  air 
or  very  low  loss  dielectric  and  must  be  insu- 
lated for  the  sum  of  the  plate  r-f  voltage  and 
the  grid  r-f  voltage. 


HANDBOOK 


Neutralizing  Procedure  255 


Figure  13 

COIL  NEUTRALIZED  AMPLIFIER 

This  neutralization  circuit  is  very  effective 
with  triode  tubes  on  any  frequency,  but  is 
particularly  effective  in  the  v-h-f  range.  The 
coil  L is  adjusted  so  that  it  resonates  at  the 
operating  frequency  with  the  grid-to-plofe 
capacitance  of  the  tube.  Capacitor  C may  be 
a very  small  unit  of  the  low-capacitance 
neutral  izing  type  and  is  used  to  trim  the  cir- 
cuit to  resonance  at  the  operating  frequency. 
If  some  means  of  varying  the  /n<fucfance  of 
the  coil  a small  amount  is  available,  the 
trimmer  capacitor  is  not  needed. 


13-6  Neutralizing 

Procedure 

An  r-f  amplifier  is  neutralized  to  prevent 
self-oscillation  or  regeneration.  A neon  bulb, 
a flashlight  lamp  and  loop  of  wire,  or  an  r-f 
galvanometer  can  be  used  as  a null  indicator 
for  neutralizing  low-power  stages.  The  plate 
voltage  lead  is  disconnected  from  the  r-f  am- 
plifier stage  while  it  is  being  neutralized. 
Normal  grid  drive  then  is  applied  to  the  r-f 
stage,  the  neutralizing  indicator  is  coupled 
to  the  plate  coil,  and  the  plate  tuning  capac- 
itor is  tuned  to  resonance.  The  neutralizing 
c^acitot  (or  capacitors)  then  can  be  adjusted 
until  minimum  r.f.  is  indicated  for  resonant 
settings  of  both  grid  and  plate  tuning  capac- 
itors. Both  neutralizing  capacitors  are  ad- 
justed simultaneously  and  to  approximately  the 
same  value  of  capacitance  when  a physically 
symmetrical  push-pull  stage  is  being  neu- 
tralized. 

A final  check  for  neutralization  should  be 
made  with  a d-c  milliammeter  connected  in  the 
grid  leak  or  grid-bias  circuit.  There  will  be 
no  movement  of  the  meter  reading  as  the  plate 
circuit  is  tuned  through  resonance  (without 
plate  voltage  being  applied)  when  the  stage 
is  completely  neutralized. 

Plate  voltage  should  be  completely  removed 

by  actually  opening  the  d-c  plate  circuit.  If 

there  is  a d-c  return  through  the  plate  supply, 
a small  amount  of  plate  current  will  flow  when 


grid  excitation  is  applied,  even  though  no  pri- 
mary a- c voltage  is  being  fed  to  the  plate  trans- 
former. 

A further  check  on  the  neutralization  of  any 
r-f  amplifier  can  be  made  by  noting  whether 
maximum  grid  current  on  the  stage  comes  at 
the  same  point  of  tuning  on  the  plate  tuning 
capacitor  as  minimum  plate  current.  This  check 
is  made  with  plate  voltage  on  the  amplifier 
and  with  normal  antenna  coupling.  As  the  plate 
tuning  capacitor  is  detuned  slightly  from  reso- 
nance on  either  side  the  grid  current  on  the 
stage  should  decrease  the  same  amount  and 
without  any  sudden  jumps  on  either  side  of 
resonance.  This  will  be  found  to  be  a very 
precise  indication  of  accurate  neutralization 
in  either  a triode  or  beam-tetrode  r-f  amplifier 
stage,  so  long  as  the  stage  is  feeding  a load 
which  presents  a resistive  impedance  at  the 
operating  frequency. 

Push-pull  circuits  usually  can  be  more  com- 
pletely neutralized  than  single-ended  circuits 
at  very  high  frequencies.  In  the  intermediate 
range  of  from  3 to  15  Me.,  single-ended  cir- 
cuits will  give  satisfactory  results. 

Neutralization  of  Radio-frequency  amplifiers 

Screen-Grid  R-F  using  screen-grid  tubes  can 

Amplifiers  be  operated  without  any  ad- 

ditional provision  for  neu- 
tralization at  frequencies  up  to  about  15  Me., 
provided  adequate  shielding  has  been  provided 
between  the  input  and  output  circuits.  Special 
v-h-f  screen-grid  and  beam  tetrode  tubes  such 
as  the  2E26,  807W,  and  5516  in  the  low-power 
category  and  HK-257B,  4E27/8001,  4-125A, 
and  4-250A  in  the  medium-power  category  can 
frequently  be  operated  at  frequencies  as  high 
as  100  Me.  without  any  additional  provision 
for  neutralization.  Tubes  such  as  the  807, 
2E22,  HY-69,  and  813  can  be  operated  with 
good  circuit  design  at  frequencies  up  to  30 
Me.  without  any  additional  provision  for  neu- 
tralization. The  815  tube  has  been  found  to 
require  neutralization  in  many  cases  above 
20  Me.,  although  the  829B  tube  will  operate 
quite  stably  at  100  Me.  without  neutralization. 

None  of  these  tubes,  however,  has  perfect 
shielding  between  the  grid  and  the  plate,  a 
condition  brought  about  by  the  inherent  in- 
ductance of  the  screen  leads  within  the  tube 
itself.  In  addition,  unless  *Vatertight”  shield- 
ing is  used  between  the  grid  and  plate  circuits 
of  the  tube  a certain  amount  of  external  leak- 
age between  the  two  circuits  is  present.  These 
difficulties  may  not  be  serious  enough  to  re- 
quire neutralization  of  the  stage  to  prevent 
oscillation,  but  in  many  instances  they  show 
up  in  terms  of  key-clicks  when  the  stage  in 

question  is  keyed,  or  as  parasitics  when  the 

stage  is  modulated.  Unless  the  designer  of  the 
equipment  can  carefully  check  the  tetrode 


256  Generation  of  R-F  Energy 


THE  RADIO 


Figure  14 

NEUTRALIZING  CIRCUITS  FOR 
BEAM  TETRODES 


A conventional  cross  neutralized  circuit  for  use  with  push-pull  beam  tetrodes  is 
shown  at  (A).  The  neutralizing  capacitors  (NO  usually  consist  of  small  plates  or 
rods  mounted  alongside  the  plate  elements  of  the  tubes.  (B)  and  (C)  show  "grid 
neutralized"  circuits  for  use  with  a single-ended  tetrode  stage  having  either  link 
coupling  or  capacitive  coupling  into  the  grid  tank.  (D)  shows  a method  of  tuning  the 
screen-lead  inductance  ta  accomplish  neutralization  in  a single-frequency  v-h-f 
tetrode  amplifier,  while  (E)  shows  a method  of  neutralization  by  increasing  the  grid- 
ta-plate  capacitance  on  a tetrode  when  the  operating  frequency  is  higher  than  that 
frequency  where  the  tetrode  is  "self-neutralized"  as  a result  of  series  resonance 
in  the  screen  lead.  Methods  (D)  and  (E)  normally  are  not  practicable  at  frequencies 
below  about  50  Me.  with  the  usual  types  of  beam  tetrode  tubes. 


Stage  tot  miscellaneous  feedback  between  the 
grid  and  plate  circuits,  and  make  the  neces- 
sary circuit  revisions  to  reduce  this  feedback 
to  an  absolute  minimum,  it  is  wise  to  neutral- 
ize the  tetrode  just  as  if  it  were  a triode  tube. 

In  most  push-pull  tetrode  amplifiers  the  sim- 
plest method  of  accomplishing  neutralization 
is  to  use  the  cross-neutralized  capacitance 
bridge  arrangement  as  normally  employed  with 
triode  tubes.  The  neutralizing  capacittuices, 
however,  must  be  very  much  smaller  than  used 
with  triode  tubes,  v^ues  of  the  order  of  0.2 
/t/tfd.  normally  being  required  with  beam  tetrode 


tubes.  This  order  of  capacitance  is  far  less 
than  can  be  obtained  with  a conventional  neu- 
tralizing capacitor  at  minimum  setting,  so  the 
neutralizing  arrangement  is  most  commonly 
made  especially  for  the  case  at  hand.  Most 
common  procedure  is  to  bring  a conductor  (con- 
nected to  the  opposite  grid)  in  the  vicinity  of 
the  plate  itself  or  of  the  plate  tuning  capacitor 
of  one  of  the  tubes.  Either  one  or  two  such 
capacitors  may  be  used,  two  being  normally 
used  on  a higher  frequency  amplifier  in  order 
to  maintain  balance  within  the  stage. 

An  example  of  this  is  shown  in  figure  14A. 


HANDBOOK 


Tetrode  Neutralization  257 


Neutralizing  A single-ended  tetrode  r-f  am- 
Single- Ended  plifier  stage  may  be  neutral- 

Tetrode  Stages  ized  in  the  same  manner  as 
illustrated  for  a push-pull 
stage  in  figure  14A,  provided  a split-stator 
tank  capacitor  is  in  use  in  the  plate  circuit. 
However,  in  the  majority  of  single-ended  tet- 
rode r-f  amplifier  stages  a single-section  ca- 
pacitor is  used  in  the  plate  tank.  Hence,  other 
neutralization  procedures  must  be  employed 
when  neutralization  is  found  necessary. 

The  circuit  shown  in  figure  14B  is  not  a 
true  neutralizing  circuit,  in  that  the  plate-to- 
grid  capacitance  is  not  balanced  out.  However, 
the  circuit  can  afford  the  equivalent  effect  by 
isolating  the  high  resonant  impedance  of  the 
grid  tank  circuit  from  the  energy  fed  back  from 
plate  to  grid.  When  NC  and  C are  adjusted  to 
bear  the  following  ratio  to  the  grid-to-plate 
capacitance  and  the  total  capacitance  from 
grid-to-gtound  in  the  output  tube: 

C Cg|( 

both  ends  of  the  grid  tank  circuit  will  be  at  the 
same  voltage  with  respect  to  ground  as  a result 
of  t-f  energy  fed  back  to  the  grid  circuit.  This 
means  that  the  impedance  from  grid  to  ground 
will  be  effectively  equal  to  the  reactance  of 
the  gtid-to-cathode  capacitance  in  parallel 
with  the  stray  grid-to-gtound  capacitance,  since 
the  high  resonant  impedance  of  the  tuned  cir- 
cuit in  the  grid  has  been  effectively  isolated 
from  the  feedback  path.  It  is  important  to  note 
that  the  effective  grid- to-ground  capacitance 
of  the  tube  being  neutralized  includes  the 
rated  grid- to- cathode  or  input  capacitance  of 
the  tube,  the  capacitance  of  the  socket,  wiring 
capacitances  and  other  strays,  but  it  does  not 
include  the  capacitances  associated  with  the 
grid  tuning  capacitor.  Also,  if  the  tube  is  be- 
ing excited  by  capacitive  coupling  from  a pre- 
ceding stage  (as  in  figure  14C),  the  effective 
grid-to-ground  capacitance  includes  the  out- 
put capacitance  of  the  preceding  stage  and 
its  associated  socket  and  wiring  capacitances. 

Cancellation  of  The  provisions  discussed  in 
Screen- Lead  the  previous  paragraphs  are 

Inductance  for  neutralization  of  the  small, 

though  still  important  at  the 
higher  frequencies,  grid-to-plate  capacitance 
of  beam-tetrode  tubes.  However,  in  the  vicinity 
of  the  upper-frequency  limit  of  each  tube  type 
the  inductance  of  the  screen  lead  of  the  tube 
becomes  of  considerable  importance.  With  a 
tube  operating  at  a frequency  where  the  in- 
ductance of  the  screen  lead  is  appreciable, 
the  screen  will  allow  a considerable  amount 
of  energy  leak-through  from  plate  to  grid  even 


though  the  socket  terminal  on  the  tube  is  care- 
fully by-passed  to  ground.  This  condition  takes 
place  even  though  the  socket  pin  is  bypassed 
since  the  reactance  of  the  screen  lead 
will  allow  a moderate  amount  of  r-f  potential 
to  appear  on  the  screen  itself  inside  the  elec- 
trode assembly  in  the  tube.  This  effect  has 
been  reduced  to  a very  low  amount  in  such 
tubes  astheHytron55l6,  andtheEimac  4X150A 
and  4X500A  but  it  is  still  quite  appreciable  in 
most  beam-tetrode  tubes. 

The  effect  of  screen-lead  inductance  on  the 
stability  of  a stage  can  be  eliminated  at  any 
particular  frequency  by  one  of  two  methods. 
These  methods  ate:  (1)  Tuning  out  the  screen- 
lead  inductance  by  series  resonatingthe  screen 
lead  inductance  with  a capacitor  to  ground. 
This  method  is  illustrated  in  figure  14D  and  is 
commonly  employed  in  commercially-built  equip- 
ment for  operation  on  a narrow  frequency  band 
in  the  range  above  about  75  Me.  The  other 
method  (2)  is  illustrated  in  figure  14E  and 
consists  in  feeding  back  additional  energy 
from  plate  to  grid  by  means  of  a small  capac- 
itor connected  between  these  two  elements. 
Note  that  this  capacitor  is  connected  in  such 
a manner  as  to  increase  the  effective  grid-to- 
plate  capacitance  of  the  tube.  This  method 
has  been  found  to  be  effective  with  807  tubes 
in  the  range  above  50  Me.  and  with  tubes  such 
as  the  4-125A  and  4-250A  in  the  vicinity  of 
their  upper  frequency  limits. 

Note  that  both  these  methods  of  stabilizing 
a beam-tetrode  v-h-f  amplifier  stage  by  can- 
cellation of  screen-lead  inductance  are  suit- 
able only  for  operation  over  a relatively  narrow 
band  of  frequencies  in  the  v-h-f  range.  At  low- 
er frequencies  both  these  expedients  for  re- 
ducing the  effects  of  screen-lead  inductance 
will  tend  to  increase  the  tendency  toward  os- 
cillation of  the  amplifier  stage. 

Neutralizing  When  a stage  cannot  be  com- 
Problems  pletely  neutralized,  the  difficulty 
usually  can  be  traced  to  one  or 
more  of  the  following  causes:  (1)  Filament 
leads  not  by-passed  to  the  common  ground  of 
that  particular  stage.  (2)  Ground  lead  from  the 
rotor  cpnnection  of  the  split-stator  tuning  ca- 
pacitor to  filament  open  or  too  long.  (3)  Neu- 
tralizing capacitors  in  a field  of  excessive 
r.f.  from  one  of  the  tuning  coils.  (4)  Electro- 
magnetic coupling  between  grid  and  plate 
coils,  or  between  plate  and  preceding  buffer 
or  oscillator  circuits.  (5)  Insufficient  shielding 
or  spacing  between  stages,  or  between  grid 
and  plate  circuits  in  compact  transmitters. 

(6)  Shielding  placed  too  close  to  plate  circuit 
coils,  causing  induced  currents  in  the  shields. 

(7)  Parasitic  oscillations  when  plate  voltage 

is  applied.  The  cure  for  the  latter  is  mainly  a 
matter  of  cut  and  try— rearrange  the  parts. 


258  Generation  of  R-F  Energy 


THE  RADIO 


Figure  15 

GROUNDED-GRID  AMPLIFIER 

This  type  of  triode  amplifier  requires  no 
neutralization,  but  can  be  used  only  with 
tubes  having  a relatively  low  plate-to-cathode 
capacitance 


change  the  length  of  grid  or  plate  or  neutraliz- 
ing leads,  insert  a parasitic  choke  in  the  grid 
lead  or  leads,  or  eliminate  the  grid  r-f  chokes 
which  may  be  the  cause  of  a low-frequency 
parasitic  (in  conjunction  with  plate  r-f  chokes). 

13-7  Grounded  Grid 

Amplifiers 

Certain  triodes  have  a grid  configuration 
and  lead  arrangement  which  results  in  very  low 
plate  to  filament  capacitance  when  the  control 
grid  is  grounded,  the  grid  acting  as  an  effec- 
tive shield  much  in  the  manner  of  the  screen 
in  a screen-grid  tube. 

By  connecting  such  a triode  in  the  circuit  of 
figure  15,  taking  the  usual  precautions  against 
stray  capacitive  and  inductive  coupling  be- 
tween input  and  output  leads  and  components, 
a stable  power  amplifier  is  realized  which  re- 
quires no  neutralization. 

At  ultra-high  frequencies,  where  it  is  diffi- 
cult to  obtain  satisfactory  neutralization  with 
conventional  triode  circuits  (particularly  when 
a wide  band  of  frequencies  is  to  be  covered), 
the  grounded-grid  arrangement  is  about  the  only 
practicable  means  of  employing  a triode  am- 
plifier. 

Because  of  the  large  amount  of  degeneration 
inherent  in  the  circuit,  considerably  more  ex- 
citation is  required  than  if  the  same  tube  were 
employed  in  a conventional  grounded-cathode 
circuit.  The  additional  power  required  to  drive 
a triode  in  a grounded-grid  amplifier  is  not 
lost,  however,  as  it  shows  up  in  the  output  cir- 
cuit and  adds  to  the  power  delivered  to  the 
load.  But  nevertheless  it  means  that  a larger 
driver  stage  is  required  for  an  amplifier  of 


I )i 


CONVENTIONAL  TRIODE  FREQUENCY 
MULTIPLIER 

Smalt  triodes  such  as  the  6C4  operate  satis- 
factorily as  frequency  multipliers,  and  can 
deliver  output  well  into  the  v-h-f  range.  Re- 
sistor R normally  will  have  a value  in  the 
vicinity  of  100,000  ohms. 


given  output,  because  a moderate  amount  of 
power  is  delivered  to  the  amplifier  load  by  the 
driver  stage  of  a grounded-grid  amplifier. 


13-8  Frequency  Multipliers 

Quartz  crystals  and  variable-frequency  os- 
cillators are  not  ordinarily  used  for  direct  con- 
trol of  the  output  of  high-frequency  transmit- 
ters. Frequency  multipliers  are  usually  em- 
ployed to  multiply  the  frequency  to  the  desired 
value.  These  multipliers  operate  on  exact  mul- 
tiples of  the  excitation  frequency;  a 3.6-Mc. 
crystal  oscillator  can  be  made  to  control  the 
output  of  a transmitter  on  7.2  or  14.4  Me.,  or 
on  28.8  Me.,  by  means  of  one  or  more  frequency 
multipliers.  When  used  at  twice  frequency, 
they  ate  often  termed  frequency  doublers.  A 
simple  doubler  circuit  is  shown  in  figure  16. 
It  consists  of  a vacuum  tube  with  its  plate  cir- 
cuit tuned  to  twice  the  frequency  of  the  grid 
driving  circuit.  This  doubler  can  be  excited 
from  a crystal  oscillator  or  another  multiplier 
or  amplifier  stage. 

Doubling  is  best  accomplished  by  operating 
the  tube  with  high  grid  bias.  The  grid  circuit 
is  driven  approximately  to  the  normal  value  of 
d-c  grid  current  through  the  r-f  choke  and  grid- 
leak  resistor,  shown  in  figure  16.  The  resist- 
ance value  generally  is  from  two  to  five  times 
as  high  as  that  used  with  the  same  tube  for 
straight  amplification.  Consequently,  the  grid 
bias  is  several  times  as  high  for  the  same 
value  of  grid  current. 

Neutralization  is  seldom  necessary  in  a 
doubler  circuit,  since  the  plate  is  tuned  to 
twice  the  frequency  of  the  grid  circuit.  The 
impedance  of  the  grid  driving  circuit  is  very 
low  at  the  doubling  frequency,  and  thus  there 
is  little  tendency  for  self-excited  oscillation. 


HANDBOOK 


Frequency  Multipliers  259 


Li 


Figure  17 


FREQUENCY  MULTIPLIER  CIRCUITS 


The  oufput  of  a triode  v~b-f  frequency  multi* 
plier  often  maybe  increased  by  neutralization 
of  the  grid-to~plate  capacitance  as  shown  at 
(A)  above.  Such  a stage  also  may  be  oper- 
ated as  a straight  amplifier  when  the  occa- 
sion demands.  A pentode  frequency  multi- 
plier is  shown  at  (B).  Conventional  power 
tetrodes  operate  satisfactorily  as  multipliers 
so  long  as  the  output  frequency  Is  below 
about  100  Me.  Above  this  frequency  special 
v-h-f  tetrodes  must  be  used  to  obtain  satis- 
factory output. 


Frequency  doublers  require  bias  of  several 
times  cutoff;  high-^  tubes  therefore  are  desir- 
able for  this  type  of  service.  Tubes  which 
have  amplification  factors  from  20  to  200  are 
suitable  for  doubler  circuits.  Tetrodes  and 
pentodes  make  excellent  doublets.  Low-p 
triodes,  having  amplification  constants  of  from 
3 to  10,  ate  not  applicable  for  doubler  service. 
In  extreme  cases  the  grid  voltage  must  be  as 
high  as  the  plate  voltage  for  efficient  doubling 
action. 

Angle  of  Flow  The  angle  of  plate  current  flow 
in  Frequency  in  a frequency  multiplier  is  a 
Multipliers  very  important  factor  in  deter- 
mining the  efficiency.  As  the 
angle  of  flow  is  decreased  for  a given  value 
of  grid  current,  the  efficiency  increases.  To 
reduce  the  angle  of  flow,  higher  grid  bias  is 
requited  so  that  the  grid  excitation  voltage 
will  exceed  the  cutoff  value  for  a shorter  por- 
tion of  the  exciting-voltage  cycle.  For  a high 
order  of  efficiency,  frequency  doublers  should 
have  an  angle  of  flow  of  90  degrees  or  less, 
triplers  60  degrees  or  less,  and  quadruplers 


^TANK  CIRCUIT  OUTPUT  VOLTAGE 


1 (CUTOFF) 


D r K N I A 

0 A A A A A. 

A Cl  ?ec»  /J  L.  M 0 Is  U\  W Y 

''J  \ ' 


”/  \ 


N(CUTOFF) { p4 -/r V 


> — EXCITATION 

\ / VOLTAGE 


Figure  18 

ILLUSTRATING  THE  ACTION  OF  A 
FREQUENCY  DOUBLER 


45  degrees  or  less.  Under  these  conditions 
the  efficiency  will  be  on  the  same  order  as 
the  reciprocal  of  the  harmonic  on  which  the 
stage  operates.  In  other  words  the  efficiency 
of  a doubler  will  be  approximately  Vi  or  50 
per  cent,  the  efficiency  of  a triplet  will  be 
approximately  % or  33  per  cent  and  that  of  a 
quadruplet  will  be  about  25  per  cent.  With  good 
stage  design  the  efficiency  can  be  somewhat 
greater  than  these  values,  but  as  the  angle 
of  flow  is  made  greater  than  these  limiting 
values,  the  efficiency  falls  off  rapidly.  The 
reason  is  apparent  from  a study  of  figure  18. 

The  pulses  ABC,  EFG,  JKL  illustrate  180- 
degree  excitation  pulses  under  Class  B opera- 
tion, the  solid  straight  line  indicating  cutoff 
bias.  If  the  bias  is  increased  by  N times,  to 
the  value  indicated  by  the  dotted  straight  line, 
and  the  excitation  increased  until  the  peak 
r-f  voltage  with  respect  to  ground  is  the  same 
as  before,  then  the  excitation  frequency  can 
be  cut  in  half  and  the  effective  excitation 
pulses  will  have  almost  the  same  shape  as 
before.  The  only  difference  is  that  every  other 
pulse  is  missing;  MNO  simply  shows  where 
the  missing  pulse  would  go.  However,  if  the 
Q of  the  plate  tank  circuit  is  high,  it  will  have 
sufficient  flywheel  effect  to  carry  over  through 
the  missing  pulse,  and  the  only  effect  will  be 
that  the  plate  input  and  r-f  output  at  optimum 
loading  drop  to  approximately  half.  As  the  in- 
put frequency  is  half  the  output  frequency,  an 
efficient  frequency  doubler  is  the  result. 

By  the  same  token,  a triplet  or  quadruplet 
can  be  analyzed,  the  triplet  skipping  two  ex- 
citation pulses  and  the  quadruplet  three.  In 
each  case  the  excitation  pulse  ideally  should 
be  short  enough  that  it  does  not  exceed  180 
degrees  at  the  output  frequency;  otherwise  the 
excitation  actually  is  bucking  the  output  over 
a portion  of  the  cycle. 

In  actual  practice,  it  is  found  uneconomical 

to  provide  sufficient  excitation  to  run  a tripler 
or  quadruplet  in  this  fashion.  Usually  the  ex- 


260  Generation  of  R-F  Energy 


THE  RADIO 


Figure  19 

PUSH-PUSH  FREQUENCY  DOUBLER 

The  output  of  a doubler  stage  may  be  materi- 
ally increased  through  the  use  of  a push-push 
circuit  such  as  illustrated  above. 


citation  pulses  will  be  at  least  90  degrees  at 
the  exciting  frequency,  with  correspondingly 
low  efficiency,  but  it  is  more  practicable  to 
accept  the  low  efficiency  and  build  up  the  out- 
put in  succeeding  amplifier  stages.  The  effi- 
ciency can  become  quite  low  before  the  power 
gain  becomes  less  than  unity. 


Push-Push  Two  tubes  can  be  connected  in 
Multipliers  parallel  to  give  twice  the  output 
of  a single-tube  doublet.  If  the 
grids  ate  driven  out  of  phase  instead  of  in 
phase,  the  tubes  then  no  longer  work  simul- 
taneously, but  rather  one  at  a time.  The  effect 
is  to  fill  'in  the  missing  pulses  (figure  18). 
Not  only  is  the  output  doubled,  but  several 
advantages  accrue  which  cannot  be  obtained 
by  straight  parallel  operation. 

Chief  among  these  is  the  effective  neutral- 
ization of  the  fundamental  and  all  odd  harmon- 
ics, an  advantage  when  spurious  emissions 
must  be  minimized.  Another  advantage  is  that 
when  the  available  excitation  is  low  and  ex- 
citation pulses  exceed  90  degrees,  the  output 
and  efficiency  will  be  greater  than  for  the 
same  tubes  connected  in  parallel. 

The  same  arrangement  may  be  used  as  a 
quadruplet,  with  considerably  better  efficiency 
than  for  straight  parallel  operation,  because 
seldom  is  it  practicable  to  supply  sufficient 
excitation  to  permit  45  degree  excitation 
pulses.  As  pointed  out  above,  the  push-push 
arrangement  exhibits  better  efficiency  than  a 
single  ended  multiplier  when  excitation  is  in- 
adequate for  ideal  multiplier  operation. 

A typical  push-push  doubler  is  illustrated 
in  figure  19.  When  high  transconductance  tubes 
are  employed,  it  is  necessary  to  employ  a 
split-stator  grid  tank  capacitor  to  prevent  self 
oscillation;  with  well  screened  tetrodes  or 
pentodes  having  medium  values  of  ttanscon- 
ductance,  a split-coil  arrangement  with  a sin- 
gle-section capacitor  may  be  employed  (the 


Figure  20 


PUSH-PULL  FREQUENCY  TRIPLER 


The  push-pull  triplet  is  advantageous  in  the 
v-h-f  range  since  circuit  balance  is  main- 
tained both  in  the  input  and  output  circuits. 
If  the  circuit  is  neutralized  it  may  be  used 
either  as  a straight  amplifier  or  as  a triplet. 
Either  triodes  or  tetrodes  may  be  used;  dual- 
unit tetrodes  such  as  the  875,  832A^  and 
829B  are  particularly  effective  in  the  v-h-f 
range. 


center  tap  of  the  grid  coil  being  by-passed  to 
ground). 

Push-Poll  Frequency  It  is  frequently  desirable 
Triplers  in  the  case  of  u-h-f  and 

v-h-f  transmitters  that 
frequency  multiplication  stages  be  balanced 
with  respect  to  ground.  Further  it  is  just  as 
easy  in  most  cases  to  multiply  the  crystal  or 
v-f-o  frequency  by  powers  of  three  rather  than 
multiplying  by  powers  of  two  as  is  frequently 
done  on  lower  frequency  transmitters.  Hence 
the  use  of  push-pull  triplers  has  become  quite 
prevalent  in  both  commercial  and  amateur 
v-h-f  and  u-h-f  transmitter  designs.  Such  stages 
are  balanced  with  respect  to  ground  and  appear 
in  construction  and  on  paper  essentially  the 
same  as  a push-pull  r-f  amplifier  stage  with 
the  exception  that  the  output  tank  circuit  is 
tuned  to  three  times  the  frequency  of  the  grid 
tank  circuit.  A circuit  for  a push-pull  triplet 
stage  is  shown  in  figure  20. 

A push-pull  triplet  stage  has  the  further 
advantage  in  amateur  work  that  it  can  also  be 
used  as  a conventional  push-pull  r-f  amplifier 
merely  by  changing  the  grid  and  plate  coils 
so  that  they  tune  to  the  same  frequency.  This 
is  of  some  advantage  in  the  case  of  operating 
the  50-Mc.  band  with  50-Mc.  excitation,  and 
then  changing  the  plate  coil  to  tune  to  144 
Me.  for  operation  of  the  stage  as  a triplet  from 
excitation  on  48  Me.  This  circuit  arrangement 
is  excellent  for  operation  with  push-pull  beam 
tetrodes  such  as  the  6360  and  829B,  although 
a pair  of  tubes  such  as  the  2E26,  or  5763  could 
just  as  well  be  used  if  proper  attention  were 
given  to  the  matter  of  screen-lead  inductance. 


HANDBOOK 


Tank  Circuits  261 


13-9  Tank  Circuit 

Capacitances 


It  is  necessary  that  the  proper  value  of  Q 
be  used  in  the  plate  tank  circuit  of  any  r-f 
amplifier.  The  following  section  has  been  de- 
voted to  a treatment  of  the  subject,  and  charts 
are  given  to  assist  the  reader  in  the  determina- 
tion of  the  proper  L/C  ratio  to  be  used  in  a 
radio-frequency  amplifier  stage. 

A Class  C amplifier  draws  plate  current  in 
the  form  of  very  distorted  pulses  of  short  dura- 
tion. Such  an  amplifier  is  always  operated  in- 
to a tuned  inductance-capacitance  or  tank  cir- 
cuit which  tends  to  smooth  out  these  pulses, 
by  its  storage  or  tank  action,  into  a sine  wave 
of  radio-frequency  output.  Any  wave-form  dis- 
tortion of  the  carrier  frequency  results  in  har- 
monic interference  in  higher-frequency  chan- 
nels. 

A Class  A r-f  amplifier  would  produce  a sine 
wave  of  radio-frequency  output  if  its  exciting 
waveform  were  also  a sine  wave.  However,  a 
Class  A amplifier  stage  converts  its  d-c  input 
to  r-f  output  by  acting  as  a variable  resistance, 
and  therefore  heats  considerably.  A Class  C 
amplifier  when  driven  hard  with  short  pulses 
at  the  peak  of  the  exciting  waveform  acts  more 
as  an  electronic  switch,  and  therefore  can  con- 
vert its  d-c  input  to  r-f  output  with  relatively 
good  efficiency.  Values  of  plate  circuit  effi- 
ciency from  65  to  85  per  cent  are  common  in 
Class  C amplifiers  operating  under  optimum 
conditions  of  excitation,  grid  bias,  and  load- 
ing. 

Tank  Circuit  Q As  Stated  before,  the  tank  cir- 
cuit of  a Class  C amplifier 
receives  energy  in  the  form  of  short  pulses  of 
plate  current  which  flow  in  the  amplifier  tube. 
But  the  tank  circuit  must  be  able  to  store 
enough  energy  so  that  it  can  deliver  a current 
essentially  sine  wave  in  form  to  the  load.  The 
ability  of  a tank  to  store  energy  in  this  man- 
ner may  be  designated  as  the  effective  Q of 
the  tank  circuit.  The  effective  circuit  Q may 
be  stated  in  any  of  several  ways,  but  essen- 
tially the  Q of  a tank  circuit  is  the  ratio  of  the 
energy  stored  to  In  times  the  energy  lost  per 
cycle.  Further,  the  energy  lost  per  cycle  mMSt, 
by  definition,  be  equal  to  the  energy  delivered 
to  the  tank  circuit  by  the  Class  C amplifier 
tube  or  tubes. 

TheQ  of  a tank  circuit  at  resonance  is  equal 
to  its  parallel  resonant  impedance  (the  reso- 
nant impedance  is  resistive  at  resonance)  di- 
vided by  the  reactance  of  either  the  capaci- 
tor or  the  inductor  which  go  to  make  up  the 

tank.  The  inductive  reactance  is  equal  to  the 

capacitive  reactance,  by  definition,  at  reso- 
nance. Hence  we  may  state: 


Figure  21 

CLASS  C AMPLIFIER  OPERATION 

Plate  current  pulses  are  shown  at  (A),  (B), 
and  (C).  The  dip  in  the  top  of  the  plate  cur- 
rent waveform  will  occur  when  the  excitation 
voltage  is  such  that  the  minimum  plate  volt- 
age dips  below  the  maximum  grid  voltage. 
A detailed  discussion  of  the  operation  of 
Class  C amplifiers  is  given  in  Chapter  Seven. 


Xc  Xl 

where  Rl  is  the  resonant  impedance  of  the 
tank  and  Xq  is  the  reactance  of  the  tank  ca- 
pacitor and  Xl  is  the  reactance  of  the  tank 
coil.  This  value  of  resonant  impedance,  Rl, 
is  the  load  which  is  presented  to  the  Class  C 
amplifier  tube  in  a single-ended  circuit  such 
as  shown  in  figure  21. 

The  value  of  load  impedance,  Rl,  which  the 
Class  C amplifier  tube  sees  may  be  obtained, 
looking  in  the  other  direction  from  the  tank 
coil,  from  a knowledge  of  the  operating  con- 
ditions on  the  Class  C tube.  This  load  imped- 
ance may  be  obtained  from  the  following  ex- 
pression, which  is  true  in  the  general  case  of 
any  Class  C amplifier: 


2 Np  Ib  Ebb 


where  the  values  in  the  equation  have  the  char- 
acteristics listed  in  the  beginning  of  Chapter  6. 

The  expression  above  is  academic,  since 
the  peak  value  of  the  fundamental  component 
of  plate  voltage  swing,  Ep„,,  is  not  ordinarily 
known  unless  a high-voltage  peaka-c  voltmeter 
is  available  for  checking.  Also,  the  decimal 

value  of  plate  circuit  efficiency  is  not  ordinari- 
ly known  with  any  degree  of  accuracy.  How- 
ever, in  a normally  operated  Class  C amplifier 


262  Generation  of  R-F  Energy 


the  radio 


Figure  22 

RELATIVE  HARMONIC  OUTPUT 
PLOTTED  AGAINST  TANK  CIRCUIT  Q 


the  plate  voltage  swing  will  be  approximately 
equal  to  0.85  to  0.9  times  the  d-c  plate  voltage 
on  the  stage,  and  the  plate  circuit  efficiency 
will  be  from  70  to  80  pet  cent  (Np  of  0.7  to 
0.8),  the  higher  values  of  efficiency  normally 
being  associated  with  the  higher  values  of 
plate  voltage  swing.  With  these  two  assump- 
tions as  to  the  normal  Class  C amplifier,  the 
expression  for  the  plate  load  impedance  can 
be  greatly  simplified  to  the  following  approxi- 
mate but  useful  expression; 


2 


which  means  simply  that  the  resistance  pre- 
sented by  the  tank  circuit  to  the  Class  C tube 
is  approximately  equal  to  one-half  the  d-c  load 
resistance  which  the  Class  C stage  presents 
to  the  power  supply  (and  also  to  the  modulator 
in  case  high-level  modulation  of  the  stage  is 
to  be  used). 

Combining  the  above  simplified  expression 
for  the  r-f  impedance  presented  by  the  tank  to 
the  tube,  with  the  expression  for  tank  Q given 
in  a previous  paragraph  we  have  the  following 
expression  which  relates  the  reactance  of  the 
tank  capacitor  or  coil  to  the  d-c  input  to  the 
Class  C stage: 

Xc  = 

2Q 

The  above  expression  is  the  basis  of  the 
usual  charts  giving  tank  capacitance  for  the 
various  bands  in  terms  of  the  d-c  plate  voltage 
and  current  to  the  Class  C stage,  including 
the  charts  of  figure  23,  figure  24  and  figure  25. 

Harmonic  Radio-  The  problem  of  harmonic 
tion  vs.  Q radiation  from  transmitters 

has  long  been  present,  but 
it  has  become  critical  only  relatively  recently 
along  with  the  extensive  occupation  of  the 
v-h-f  range.  Television  signals  are  particularly 
susceptible  to  interference  from  other  signals 
falling  within  the  pass  band  of  the  receiver, 
so  that  theTVl  problem  has  received  the  major 
emphasis  of  all  the  services  in  the  v-h-f  range 
which  are  susceptible  to  interference  from 
harmonics  of  signals  in  the  h-f  or  lower  v-h-f 
range. 


TOTAL  CAPACITANCE  ACROSS  LC  CIRCUIT  (Cl) 


Figure  23 

PLATE-TANK  CIRCUIT  ARRANGEMENTS 

Shown  above  In  the  case  of  each  of  the  tank  circuit  types  is  the  recommended  tank  circuit  ca- 
pacitance. (A)  is  a conventional  tetrode  amplifier,  (Bl  is  a coil-neutralized  triode  amplifier, 
(C)  is  a grounded-grid  triode  amplifier,  (D)  is  a grid-neutralized  triode  amplifier. 


HANDBOOK 


Tank  Circuits  263 


CORRECT  VALUES  OF  TANK  CIRCUIT  CAPACITANCE  (C)  FOR 
OPERATING  Q OF  12  WITH  SINGLE-ENDED  SPLIT  TANK  COILS 


Figure  24 

PLATE-TANK  CIRCUIT  ARRANGEMENTS 


Shown  above  for  each  of  the  tank  circuit  types  is  the  recommended  tank  circuit  capacitance  at 
the  operating  frequency  for  an  operating  Q of  12.  (A)  is  a split~stator  tank,  each  section  of  which 
is  twice  the  capacity  value  read  on  the  graph.  {Bl  is  circuit  using  tapped  coil  for  phase  reversal. 


Inspection  of  figure  22  will  show  quickly 
that  the  tank  circuit  of  a Class  C amplifier 
should  have-an  operating  Q of  12  or  greater 
to  afford  satisfactory  rejection  of  second  har- 
monic energy.  The  curve  begins  to  straighten 
out  above  a Q of  about  15,  so  that  a consider- 
able increase  in  Q must  be  made  before  an  ap- 
preciable reduction  in  second-harmonic  energy 
is  obtained.  Above  a circuit  Q of  about  10  any 
increase  will  not  afford  appreciable  reduction 
in  the  third-harmonic  energy,  so  that  additional 
harmonic  filtering  circuits  external  to  the  am- 
plifier proper  must  be  used  if  increased  atten- 
uation of  higher  order  harmonics  is  desired. 
The  curves  also  show  that  push-pull  amplifiers 
may  be  operated  at  Q values  of  6 or  so,  since 
the  second  harmonic  is  cancelled  to  a large 
extent  if  there  is  no  unbalanced  coupling  be- 
tween the  output  tank  circuit  and  the  antenna 
system. 

Capacity  Charts  for  Figures  23,  24  and  25  il- 
Carrect  Tank  Q lustrate  the  correct  value 

of  tank  capacity  for  vari- 
ous circuit  configurations.  A Q value  of  12 
has  been  chosen  as  optimum  for  single  ended 
circuits,  and  a value  of  6 has  been  chosen  for 
push-pull  circuits.  Figure  23  is  used  when  a 
single  ended  stage  is  employed,  and  the  ca- 
pacitance values  given  are  for  the  total  ca- 
pacitance across  the  tank  coil.  This  value  in- 
cludes the  tube  interelectrode  capacitance 

(plate  to  ground),  coil  distributed  capacitance, 
wiring  capacities,  and  the  value  of  any  low- 


inductance  plate-to-ground  by-pass  capacitor 
as  used  for  reducing  harmonic  generation,  in 
addition  to  the  actual  "in-use”  capacitance 
of  the  plate  tuning  capacitor.  Total  circuit 
stray  capacitance  may  vary  from  perhaps  5 
micromicrofarads  for  a v-h-f  stage  to  30  micro- 
microfarads  for  a medium  power  tetrode  h-f 
stage. 

When  a split  plate  tank  coil  is  employed  in 
the  stage  in  question,  the  graph  of  figure  24 
should  be  used.  The  capacity  read  from  the 
graph  is  the  total  capacity  across  the  tank 
coil.  If  the  split-stator  tuning  capacitor  is 
used,  each  section  of  the  capacitor  should 
have  a value  of  capacity  equal  to  twice  the 
value  indicated  by  the  graph.  As  in  the  case 
of  figure  23,  the  values  of  capacity  read  on 
the  graph  of  figure  24  include  all  residual  cir- 
cuit capacities. 

For  push-pull  operation,  the  correct  values 
of  tank  circuit  capacity  may  be  determined 
with  the  aid  of  figure  25.  The  capacity  values 
obtained  from  figure  25  are  the  effective  values 
across  the  tank  circuit,  and  if  a split-stator 
tuning  capacitor  is  used,  each  section  of  the 
capacitor  should  have  a value  of  capacity  e- 
qual  to  twice  the  value  indicated  by  the  graph. 
As  in  the  case  of  figures  23  and  24,  the  values 
of  capacity  read  on  the  graph  of  figure  25  in- 
clude all  residual  circuit  capacities. 

The  tank  circuit  operates  in  the  same  man- 

ner whether  the  tube  feeding  it  is  a pentode, 

beam  tetrode,  neutralized  triode,  grounded- 
grid  triode,  whether  it  is  single  ended  or  push- 


264  Generation  of  R-F  Energy 


THE  RADIO 


CORRECT  VALUES  OF  TANK  CIRCUIT  CAPACITANCE  (Cl)  FOR 
OPERATING  Q OF  6 WITH  PUSH-PULL  TANK  CIRCUITS 


Figure  25 

PLATE-TANK  CIRCUIT  ARRANGEMENTS  FOR  PUSH-PULL  STAGES 

Shown  above  is  recommended  tank  circuit  capacity  at  operating  frequency  for  a Q of  6.  (A)  iff 
spl it-statar  tank^  each  section  of  which  is  twice  the  capacity  value  read  on  the  graph.  (B)  is 
circuit  using  tapped  coil  for  phase  reversal. 


pull,  or  whether  it  is  shunt  fed  or  series  fed. 
The  important  thing  in  establishing  the  oper- 
ating Q of  the  tank  circuit  is  the  ratio  of  the 
loaded  resonant  impedance  across  its  termi- 
nals ro  the  reactance  of  the  L and  the  C which 
make  up  the  tank. 

Due  to  the  unknowns  involved  in  determin- 
ing circuit  stray  capacitances  it  is  sometimes 
more  convenient  to  determine  the  value  of  L 
requited  for  the  proper  circuit Q (by  the  method 
discussed  earlier  in  this  Section)  and  then  to 
vary  the  tuned  circuit  capacitance  until  reso- 
nance is  reached.  This  method  is  most  fre- 
quently used  in  obtaining  proper  circuit  Q in 
commercial  transmitters. 

The  values  of  Rp  for  using  the  charts  are 
easily  calculated  by  dividing  the  d-c  plate  sup- 
ply voltage  by  the  total  d-c  plate  current  (ex- 
pressed in  amperes).  Correct  values  of  total 
tuning  capacitance  are  shown  in  the  chart  for 
the  different  amateur  bands.  Tbe  shunt  stray 
capacitance  can  be  estimated  closely  enough 
for  all  practical  purposes.  The  coil  inductance 
should  then  be  chosen  which  will  produce 
resonance  at  the  desired  frequency  with  the 
total  calculated  tuning  capacitance. 

Effect  of  Load-  The  Q of  a circuit  depends 
ing  on  Q upon  the  resistance  in  series 

with  the  capacitance  and  in- 
ductance. This  series  resistance  is  very  low 
for  a low-loss  coil  not  loaded  by  an  antenna 
circuit.  The  value  of  Q may  be  from  100  to  600 
under  these  conditions.  Coupling  an  antenna 


circuit  has  the  effect  of  increasing  the  series 
resistance,  though  in  this  case  the  power  is 
consumed  as  useful  radiation  by  the  antenna. 
Mathematically,  the  antenna  increases  the 
value  of  R in  the  expression  Q = cuL/R  where 
L is  the  coil  inductance  in  microhenry  s and 
<u  is  the  term  27rf,  f being  in  megacycles. 

The  coupling  from  the  final  tank  circuit  to 
the  antenna  or  antenna  transmission  line  crui 
be  varied  to  obtain  values  of  Q from  perhaps 
3 at  maximum  coupling  to  a value  of  Q equal 
to  the  unloaded  Q of  the  circuit  at  zero  an- 
tenna coupling.  This  value  of  unloaded  Q can 
be  as  high  as  500  or  600,  as  mentioned  in  the 
preceding  paragraph.  However,  the  value  of 
Q = 12  will  not  be  obtained  at  values  of  nor- 
mal d-c  plate  current  in  the  Class  C amplifier 
stage  unless  the  C-to-L  ratio  in  the  tank  cir- 
cuit is  correct  for  that  frequency  of  operation. 

Tuning  Capacitor  To  determine  the  requited 
Air  Gap  tuning  capacitor  ait  gap  for 

a particular  amplifier  cir- 
cuit it  is  first  necessary  to  estimate  the  peak 
r-f  voltage  which  will  appear  between  the 
plates  of  the  tuning  capacitor.  Then,  using 
figure  26,  it  is  possible  to  estimate  the  plate 
spacing  which  will  be  requited. 

The  instantaneous  r-f  voltage  in  the  plate 
circuit  of  a Class  C amplifier  tube  varies  from 
neatly  zero  to  nearly  twice  the  d-c  plate  volt- 
age. If  the  d-c  voltage  is  being  100  pet  cent 
modulated  by  an  audio  voltage,  the  t-f  peaks 
will  reach  nearly  four  times  the  d-c  voltage. 


HANDBOOK 


L and  Pi  Networks  265 


FIGURE  26 


USUAL  BREAKDOWN  RATINGS  OF 
COMMON  PLATE  SPACINGS 
Air-gap  in  Peak  voltage 

inches  breokdown 


,030  1,000 

.050  2,000 

.070  3,000 

.100  4,000 

.125  4.500 

.150  5,200 

.170  6,000 

.200  7,500 

.250  9,000 

.350  11,000 

.500  15,000 

.700  20,000 


Recommended  air-gap  for  use  when  no  d-c 
voltage  appears  across  plate  tank  condenser 
(when  plate  circuit  is  shunt  fed,  or  when  the 
plate  tank  condenser  is  insulated  from 
ground). 


D.C.  PLATE 
VOLTAGE 

C.W. 

1 PLATE 

1 MOD. 

400 

.030 

.050 

600 

.050 

,070 

750 

.050 

.084 

1000 

.070 

.100 

1250 

.070 

.144 

1500 

.078 

.200 

2000 

.100 

.250 

2500 

.175 

.375 

3000 

.200 

.500 

3500 

.250 

.600 

Spacings  should  be  multiplied  by  1.5  for 
same  safety  factor  when  d-c  voltage  appears 
across  plate  tank  condenser. 


These  rules  apply  to  a loaded  amplifier  or 
buffer  stage.  If  either  is  operated  without  an 
r-f  load,  the  peak  voltages  will  be  greater  and 
can  exceed  the  d-c  plate  supply  voltage.  For 
this  reason  no  amplifier  should  be  operated 
without  load  when  anywhere  near  normal  d-c 
plate  voltage  is  applied. 

If  a plate  blocking  condenser  is  used,  it 
must  be  rated  to  withstand  the  d-c  plate  volt- 
age plus  any  audio  voltage.  This  capacitor 
should  be  rated  at  a d-c  working  voltage  of  at 
ItdLSt  twice  the  d-c  plate  supply  in  a plate  mod- 
ulated amplifier,  and  at  least  equal  to  the  d-c 
supply  in  any  other  type  of  r-f  amplifier. 


13-10  L and  Pi  Matching 
Networks 


The  L and  pi  networks  often  can  be  put  to 
advantageous  use  in  accomplishing  an  imped- 
ance match  between  two  differing  impedances. 
Common  applications  are  the  matching  between 
a transmission  line  and  an  antenna,  or  between 
the  plate  circuit  of  a single-ended  amplifier 
stage  and  an  antenna  transmission  line.  Such 
networks  may  be  used  to  accomplish  a match 


Rp=  Ra(Q2-|-i)Cexact) 


Rp  = Q2  Ra  (approx.) 


o - ^ = 2^  = E£_  - E£. 
Ra  Ra  Xc  Xl 

Xl=Xc 


^RP  = 
Rp  = 


APPROX,  • 
225  Ra 


PLATE  VOLTAGE 
PLATE  CURRENT 


Xr  = 

TT 

Xl  = 

C 15 


Figure  27 

THE  L NETWORK  IMPEDANCE 
TRANSFORMER 

The  L network  is  useful  with  a moderate 
operating  Q for  high  values  of  impedance 
transformation,  and  it  may  be  used  for  appli- 
cations other  than  in  the  plate  circuit  of  a 
tube  with  relatively  low  values  of  operating 
Q for  moderate  impedance  transformations. 
Exact  and  approximate  design  equations  are 
given. 


between  the  plate  tank  circuit  of  an  amplifier 
and  a transmission  line,  or  they  may  be  used 
to  match  directly  from  the  plate  circuit  of  an 
amplifier  to  the  line  without  the  requirement 
for  a tank  circuit— provided  the  network  is  de- 
signed in  such  a manner  that  it  has  sufficient 
operating  Q for  accomplishing  harmonic  atten- 
uation. 

The  L Matching  The  L network  is  of  limited 
Network  Utility  in  impedance  match- 

ing since  its  ratio  of  imped- 
ance transformation  is  fixed  at  a value  equal 
to  (Q^+1).  The  operating  Q may  be  relatively 
low  (perhaps  3 to  6)  in  a matching  network  be- 
tween the  plate  tank  circuit  of  an  amplifier 
and  a transmission  line;  hence  impedance 
transformation  ratios  of  10  to  1 and  even  lower 
may  be  attained.  But  when  the  network  also 
acts  as  the  plate  tank  circuit  of  the  amplifier 
stage,  as  in  figure  27,  the  operating  Q should 
be  at  least  12  and  preferably  15.  An  operating 
Q of  15  represents  an  impedance  transforma- 
tion of  225;  this  value  normally  will  be  too 
high  even  for  transforming  from  the  2000  to 
10,000  ohm  plate  impedance  of  a Class  C am- 
plifier stage  down  to  a 50-ohm  transmission 
line. 

However,  the  L network  is  interesting  since 
it  forms  the  basis  of  design  for  the  pi  network. 
Inspection  of  figure  27  will  show  that  the  L 
network  in  reality  must  be  considered  as  a 
parallel-resonant  tank  circuit  in  which 
represents  the  coupled-in  load  resistance; 

only  in  this  case  the  load  resistance  is  di- 
rectly coupled  into  the  tank  circuit  rather  than 
being  inductively  coupled  as  in  the  conven- 


266  Generation  of  R-F  Ener 


THE  RADIO 


gy 


Ebb 

Ib 


Rd.c,  = 

XCl 
Xli 


2 

Rp 

Q 

Rp 


Xc2  =-Ra 


Rp 


XL2  = 


Ra(QZ  + i)-Rp 
R*=-  Xca 


Ra2  + Xc2=' 
Xltot.  = X|_,  + Xl2 


Figure  28 

THE  PI  NETWORK 

T/ie  pi  network  is  valuable  for  use  as  an  int~ 
pedance  transformer  over  a wide  ratio  of 
transformation  values*  The  operating  Q should 
be  at  least  12  and  preferably  15  to  20  when 
the  circuit  is  to  be  used  in  the  plate  circuit 
of  a Class  C amplifier.  Design  equations 
are  given  above.  The  inductor  Ltot  tepre- 
sents  a single  inductance,  usually  variable, 
with  a value  equal  to  the  sum  of  Lj  and  Lj. 


tional  arrangement  where  the  load  circuit  is 
coupled  to  the  tank  circuit  by  means  of  a link. 
When  is  shorted,  L and  C comprise  a con- 
ventional parallel-resonant  tank  circuit,  since 
for  proper  operation  L and  C must  be  resonant 
in  order  for  the  network  to  present  a resistive 
load  to  the  Class  C amplifier. 


The  Pi  Network  The  pi  impedance  matching 
network,  illustrated  in  figure 
28,  is  much  more  general  in  its  application 
than  the  L network  since  it  offers  greater  har- 
monic attenuation,  and  since  it  can  be  used 
to  match  a relatively  wide  range  of  impedances 
while  still  maintaining  any  desired  operating 
Q.  The  values  of  C,  and  L,  in  the  pi  network 
of  figure  28  can  be  thought  of  as  having  the 
same  values  of  the  L network  in  figure  27  for 
the  same  operating  Q,  but  what  is  more  impor- 
tant from  the  comparison  standpoint  these  val- 
ues will  be  the  same  as  in  a conventional  tank 
circuit. 

The  value  of  the  capacitance  may  be  deter- 
mined by  calculation,  with  the  operating  Q and 
the  load  impedance  which  should  be  reflected 
to  the  plate  of  the  Class  C amplifier  as  the 
two  knowns— or  the  actual  values  of  the  ca- 


pacitance may  be  obtained  for  an  operating  Q 
of  12  by  reference  to  figures  23,  24  and  25. 

The  inducrive  arm  in  the  pi  network  can  be 
thought  of  as  consisting  of  two  inductances 
in  series,  as  illustrated  in  figure  28.  The  first 
portion  of  this  inductance,  Li,  is  that  value  of 
inductance  which  would  resonate  with  C,  at 
the  operating  frequency— the  same  as  in  a con- 
ventional tank  circuit.  However,  the  actual 
value  of  inductance  in  this  atm  of  the  pi  net- 
work, L,ot  will  be  greater  than  L,  for  normal 
values  of  impedance  transformation.  For  high 
transformation  ratios  Lj*  will  be  only  slightly 
greater  than  L,;  for  a transformation  ratio  of 
1.0,  L,o,  will  be  twice  as  great  as  L,.  The 
amount  of  inductance  which  must  be  added  to 
L,  to  restore  resonance  and  maintain  circuit 
Q is  obtained  through  use  of  the  expression 
for  X„j  in  figure  28. 

The  peak  voltage  rating  of  the  main  tuning 
capacitor  C,  should  be  the  normal  value  for  a 
Class  C amplifier  operating  at  the  plate  volt- 
age to  be  employed.  The  inductor  Ltot  uiay  be 
a plug-in  coil  which  is  changed  for  each  band 
of  operation,  or  some  sort  of  variable  inductor 
may  be  used.  A continuously  variable  slider- 
type  of  variable  inductor,  such  as  used  in  cer- 
tain items  of  surplus  military  equipment,  may 
be  used  to  good  advantage  if  available,  or  a 
ta^ed  inductor  such  as  used  in  the  ART-13 
may  be  employed.  However,  to  maintain  good 
circuit  Q on  the  higher  frequencies  when  a 
variable  or  tapped  coil  is  used  on  the  lower 
frequencies,  the  tapped  or  variable  coil  should 
be  removed  from  the  circuit  and  replaced  by 
a smaller  coil  which  has  been  especially  de- 
signed for  the  higher  frequency  ranges. 

The  peak  voltage  rating  of  the  output  or 
loading  capacitor,  Cj,  is  determined  by  the 
power  level  and  the  impedance  to  be  fed.  If  a 
50-ohm  coaxial  line  is  to  be  fed  from  the  pi 
network,  receiving-type  capacitors  will  be 
satisfactory  even  up  to  the  power  level  of  a 
plate-modulated  kilowatt  amplifier.  In  any 
event,  the  peak  voltage  which  will  be  im- 
pressed across  the  output  capacitor  is  ex- 
pressed by:  Epk*  = 2 R^  Wp,  where  Epj;  is  the 
peak  voltage  across  the  capacitor,  Ra  is  the 
value  of  resistive  load  which  the  network  is 
feeding,  and  Wp  is  the  maximum  value  of  the 
average  power  output  of  the  stage.  The  har- 
monic attenuation  of  the  pi  network  is  quite 
good,  although  an  external  low-pass  filter  will 
be  required  to  obtain  harmonic  attenuation 
value  upward  of  100  db  such  as  normally  re- 
quired. The  attenuation  to  second  harmonic 
energy  will  be  approximately  40  db  for  an  oper- 
ating Q of  15  for  the  pi  network;  the  value 
increases  to  about  45  db  for  a 1:1  transforma- 
tion and  falls  to  about  38  db  for  an  impedance 
step-down  of  80:1,  assuming  that  the  oper- 
ating Q is  maintained  at  15. 


HANDBOOK 


Grid  Bias  267 


c-  f,  WHERE  EB  is  plate  VOL  TASE 

PLATE  LOAD  (ohms)  — and  Ib  /5  plate  current 
in  V^MPEfKES . 

Ca  - .0002$  JJF.  MICA  capacitor  rated  at  twice  the  d.c. 

PLATE  VOLTAGE. 

RFC  I 26  ENAMELED.  CLOSE-WOUND  ON  A CERAMIC  INSULATOR 
l-OIA.,  A'-LONG  OR  NATIONAL  R-I7SA 

RFC 2-  2j  MH,  NATIONAL  R-lOO 


Estimated  Plate 
Load  (ohms) 

1,000 

1,500 

2,000 

2,500 

3,000 

3,500 

4,000 

4,500 

5,000 

6,000* 

NOTES 

Cl  in  3.5  Me 

520 

360 

280 

210 

180 

155 

135 

120 

110 

90 

The  actual  capacitance  setting 

7 

260 

180 

140 

105 

90 

76 

68 

60 

56 

45 

for  Cl  equals  the  value  in  this 

14 

130 

90 

70 

52 

45 

38 

34 

30 

28 

23 

table  minus  the  published  tube 

21 

85 

60 

47 

35 

31 

25 

23 

20 

19 

15 

output  capacitance.  Air  gap 

28 

65 

45 

35 

26 

23 

19 

17 

15 

14 

11 

approx.  10  mils/lOO  v Eh. 

4.5 

6.5 

8.5 

10.5 

12.5 

14 

15.5 

18 

20 

25 

Inductance  values  are  for  a 

7 

2.2 

3.2 

4.2 

5.2 

6.2 

7 

7.8 

9 

10 

12.5 

50-ohm  load.  For  a 70-ohm 

14 

1.1 

1.6 

2.1 

2.6 

3.1 

3.5 

3.9 

4.5 

5 

6.2 

load,  values  are  approx.  3% 

21 

0.73 

1.08 

1.38 

1.7 

2.05 

2.3 

2.6 

3 

3.3 

4.1 

higher. 

28 

0.55 

0.8 

1.05 

1.28 

1.55 

1.7 

1.95 

2.25 

2.5 

3.1 

in  Miuf,  3.5  Me 

2,400 

2,100 

1,800 

1,550 

1,400 

1,250 

1,100 

1,000 

900 

700 

For  50>ohm  transmission  line. 

7 

1,200 

1,060 

900 

760 

700 

630 

560 

500 

460 

350 

Air  gap  for  Ce  is  approx.  1 

14 

600 

530 

450 

380 

350 

320 

280 

250 

230 

175 

mil/iOO  V E,,. 

21 

400 

350 

300 

250 

230 

210 

185 

165 

155 

120 

28 

300 

265 

225 

190 

175 

160 

140 

125 

115 

90 

Ca  in  3.5  Me 

1,800 

1,500 

1,300 

1,100 

1,000 

900 

800 

720 

640 

500 

For  70>ohm  transmission  line. 

7 

900 

750 

650 

560 

500 

450 

400 

360 

320 

250 

14 

450 

370 

320 

280 

250 

220 

200 

180 

160 

125 

21 

300 

250 

215 

190 

170 

145 

130 

120 

no 

85 

28 

225 

185 

160 

140 

125 

110 

100 

90 

80 

65 

* Values  given  are  approximations.  All  components  shown  in  Table  I are  for  o Q of  12.  For  other  values  of  Q,  use 


Oh  Ca  Qa 

— — . and  

Q„  C,,  Q„ 


It. 

Lh 


When  the  estimated  plate  load  is  higher  than  5,000  ohms,  it  is  recommended  that  the 


components  be  selected  for  a circuit  Q between  20  end  30. 


Table  I Components  for  Pi-Coupled  Final  Amplifiers 


Component  Chart  To  simplify  design  pro- 
for  Pi-Networks  cedure,  a pi-network  chart, 
compiled  by  M.  Seybold, 
W2RYI  (reproduced  by  courtesy  of  R.C.A.  Tube 
Division,  Harrison,  N.J.)  is  shown  in  table  1. 
This  chart  summarizes  the  calculations  of  fig- 
ure 28  for  various  values  of  plate  load. 


13-1 1 Grid  Bias 

Radio-frequency  amplifiers  require  some 
form  of  grid  bias  lor  proper  operation.  Prac- 
tically all  r-f  amplifiers  operate  in  such  a man- 
ner that  plate  current  flows  in  the  form  of 

short  pulses  which  have  a duration  of  only  a 
fraction  of  an  r-f  cycle.  To  accomplish  this 


with  a sinusoidal  excitation  voltage,  the  oper- 
ating grid  bias  must  be  at  least  sufficient  to 
cut  off  the  plate  current.  In  very  high  efficien- 
cy Class  C amplifiers  the  operating  bias  may 
be  many  times  the  cutoff  value.  Cutoff  bias, 
it  will  be  recalled,  is  that  value  of  grid  volt- 
age which  will  reduce  the  plate  current  to 
zero  at  the  plate  voltage  employed.  The  method 
for  calculating  it  has  been  indicated  previous- 
ly. This  theoretical  value  of  cutoff  will  not 
reduce  the  plate  current  completely  to  zero, 
due  to  the  variable-p  tendency  or  "knee” 
which  is  characteristic  of  all  tubes  as  the 
cutoff  point  is  approached. 

Class  C Bios  Amplitude  modulated  Class  C 

amplifiers  should  be  operated 

with  the  grid  bias  adjusted  to  a value  greater 
than  twice  cutoff  at  the  operating  plate  volt- 


268  Generation  of  R-F  Energy 


THE  RADIO 


Figure  29 

GRID-LEAK  BIAS 

The  grid  leak  on  an  amplifier  or  multiplier 
stage  may  also  be  used  as  the  shunt  feed 
impedance  to  the  grid  of  the  tube  when  a 
high  value  of  grid  leak  (greater  than  perhaps 
20,000  ohms)  is  used.  When  a lower  value  of 
grid  leak  is  to  be  employed,  an  r~f  choke 
should  be  used  between  the  grid  of  the  tube 
and  the  grid  leak  to  reduce  r-f  losses  in  the 
grid  leak  resistance. 


age.  This  procedure  will  insure  that  the  tube 
is  operating  at  a bias  greater  than  cutoff  when 
the  plate  voltage  is  doubled  on  positive  modu- 
lation peaks.  C-w  telegraph  and  FM  trans- 
mitters can  be  operated  with  bias  as  low  as 
cutoff,  if  only  limited  excitation  is  available 
and  moderate  plate  efficiency  is  satisfactory. 
In  a c-w  transmitter,  the  bias  supply  or  re- 
sistor should  be  adjusted  to  the  point  which 
will  allow  normal  grid  current  to  flow  for  the 
particular  amount  of  grid  driving  r-f  power 
available.  This  form  of  adjustment  will  allow 
more  output  from  the  under-excited  t-f  ampli- 
fier than  when  higher  bias  is  used  with  corre- 
sponding lower  values  of  grid  current.  In  any 
event,  the  operating  bias  should  be  set  at  as 
low  a value  as  will  give  satisfactory  opera- 
tion, since  harmonic  generation  in  a stage  in- 
creases rapidly  as  the  bias  is  increased. 

Grid-Leak  Bias  A resistor  can  be  connected 
in  the  grid  circuit  of  a Class 
C amplifier  to  provide  grid-leak  bias.  This  re- 
sistor, Ri  in  figure  29,  is  part  of  the  d-c  path 
in  the  grid  circuit. 

The  r-f  excitation  applied  to  the  grid  cir- 
cuit of  the  tube  causes  a pulsating  direct  cur- 
rent to  flow  through  the  bias  supply  lead,  due 
to  the  rectifying  action  of  the  grid,  and  any 
current  flowing  through  R,  produces  a voltage 
drop  across  that  resistor.  The  grid  of  the  tube 
is  positive  for  a short  duration  of  each  r-f 
cycle)  and  draws  electrons  from  the  filament 
or  cathode  of  the  tube  during  that  time.  These 
electrons  complete  the  circuit  through  the  d-c 
grid  return.  The  voltage  drop  across  the  re- 
sistance in  the  grid  return  provides  a nega- 
tive bias  for  the  grid. 

Grid-leak  bias  automatically  adjusts  itself 
over  fairly  wide  variations  of  r-f  excitation. 
The  value  of  grid-leak  resistance  should  be 
such  that  normal  values  of  grid  current  will 
flow  at  the  maximum  available  amount  of  r-f 


-BIAS  SUPPLY 


Figure  30 

COMBINATION  GRID-LEAK  AND 
FIXED  BIAS 

Grid-leak  bias  often  is  used  in  conjunction 
with  a fixed  minimum  value  of  power  supply 
bias.  This  arrangement  permits  the  operating 
bias  to  be  established  by  the  excitation  ener- 
gy, but  in  the  absence  of  excitation  the  elec- 
trode currents  to  the  tube  will  be  held  to  safe 
values  by  the  fixed-minimum  power  supply 
bias.  If  a relatively  low  value  of  grid  leak 
is  to  be  used,  an  r-f  choke  should  be  con- 
nected between  the  grid  of  the  tube  and  the 
grid  leak  os  discussed  in  figure  29, 


excitation.  Grid-leak  bias  cannot  be  used  for 
grid-modulated  or  linear  amplifiers  in  which 
the  average  d-c  grid  current  is  constantly 
varying  with  modulation. 


Sofety  Bias  Grid-leak  bias  alone  provides  no 
protection  against  excessive 
plate  current  in  case  of  failure  of  the  source 
of  r-f  grid  excitation.  A C-battery  or  C-bias 
supply  can  be  connected  in  series  with  the 
grid  leak,  as  shown  in  figure  30.  This  fixed 
protective  bias  will  protect  the  tube  in  the 
event  of  failure  of  grid  excitation.  "Zero-bias” 
tubes  do  not  require  this  bias  source  in  addi- 
tion to  the  grid  leak,  since  their  plate  current 
will  drop  to  a safe  value  when  the  excitation 
is  removed. 

Cathode  Bias  A resistor  can  be  connected  in 
series  with  the  cathode  or  cen- 
ter-tapped filament  lead  of  an  amplifier  to  se- 
cure automatic  bias.  The  plate  current  flows 
through  this  resistor,  then  back  to  the  cathode 
<x*  filament,  and  the  voltage  drop  across  the 
resistor  can  be  applied  to  the  grid  circuit  by 
connecting  the  grid  bias  lead  to  the  grounded 
or  power  supply  end  of  the  resistor  R,  as  shown 
in  figure  31. 

The  grounded  (B-minus)  end  of  the  cathode 
resistor  is  negative  relative  to  the  cathode 
by  an  amount  equal  to  the  voltage  drop  across 
the  resistor.  The  value  of  resistance  must  be 
so  chosen  that  the  sum  of  the  desired  grid 
and  plate  current  flowing  through  the  resistor 
will  bias  the  tube  for  proper  operation. 

This  type  of  bias  is  used  more  extensively 
in  audio-frequency  than  in  radio-frequency  am- 
plifiers. The  voltage  drop  across  the  resistor 


HANDBOOK 


Protective  Circuits  269 


BATTERY  - 


Figure  31 

R-F  STAGE  WITH  CATHODE  BIAS 

Cathode  bias  sometimes  is  advantageous  for 
use  in  an  r-f  stage  that  operates  with  a refa- 
tively  small  amount  of  r-f  excitation. 


Figure  32 

R-F  STAGE  WITH  BATTERY  BIAS 

Battery  bias  is  seldom  used,  due  to  c/efer/oro- 
tion  of  the  cells  by  the  reverse  grid  current. 
However,  it  may  be  used  in  certain  special 
applications,  or  the  fixed  bias  voltage  may 
be  supplied  by  a bias  power  supply. 


must  be  subtracted  from  the  total  plate  supply 
voltage  when  calculating  the  power  input  to 
the  amplifier,  and  this  loss  of  plate  voltage 
in  an  r-f  amplifier  may  be  excessive.  A Class 
A audio  amplifier  is  biased  only  to  approxi- 
mately one-half  cutoff,  whereas  an  r-f  amplifier 
may  be  biased  to  twice  cutoff,  or  more,  and 
thus  the  plate  supply  voltage  loss  may  be  a 
large  percentage  of  the  total  available  voltage 
when  using  low  or  medium  (i  tubes. 

Oftentimes  just  enough  cathode  bias  is  em- 
ployed in  an  r-f  amplifier  to  act  as  safety  bias 
to  protect  the  tubes  in  case  of  excitation  fail- 
ure, with  the  rest  of  the  bias  coming  from  a 
grid  leak. 

Separate  Bias  An  external  supply  often  is 
Supply  used  for  grid  bias,  as  shown  in 

figure  32.  Battery  bias  gives 
very  good  voltage  regulation  and  is  satisfac- 
tory for  grid-modulated  or  linear  amplifiers, 
which  operate  at  low  grid  current.  In  the  case 
of  Class  C amplifiers  which  operate  with  high 
grid  current,  battery  bias  is  not  satisfactory. 
This  direct  current  has  a charging  effect  on 
the  dry  batteries;  after  a few  months  of  service 
the  cells  will  become  unstable,  bloated,  and 
noisy. 

A separate  a-c  operated  power  supply  is 
commonly  used  for  grid  bias.  The  bleeder  re- 
sistance across  the  output  of  the  filter  can  be 
made  sufficiently  low  in  value  that  the  grid 
current  of  the  amplifier  will  not  appreciably 
change  the  amount  of  negative  grid-bias  volt- 
age. Alternately,  a voltage  regulated  grid-bias 
supply  can  be  used.  This  type  of  bias  supply 
is  used  in  Class  B audio  and  Class  B r-f  lin- 
ear amplifier  service  where  the  voltage  regu- 
lation in  the  C-bias  supply  is  important.  For 
a Class  C amplifier,  regulation  is  not  so  im- 
portant, and  an  economical  design  of  compo- 
nents in  the  power  supply,  therefore,  can  be 
utilized.  In  this  case,  the  bias  voltage  must 
be  adjusted  with  normal  grid  current  flowing, 
as  the  grid  current  will  raise  the  bias  con- 


siderably when  it  is  flowing  through  the  bias- 
supply  bleeder  resistance. 


13-12  Protective  Circuits  for 
Tetrode  Transmitting  Tubes 

The  tetrode  transmitting  tube  requires  three 
operating  voltages:  grid  bias,  screen  voltage, 
and  plate  voltage.  The  current  requirements  of 
these  three  operating  voltages  are  somewhat 
interdependent,  and  a change  in  potential  of 
one  voltage  will  affect  the  current  drain  of  the 
tetrode  in  respect  to  the  other  two  voltages. 
In  particular,  if  the  grid  excitation  voltage  is 
interrupted  as  by  keying  action,  or  if  the  plate 
supply  is  momentarily  interrupted,  the  resulting 
voltage  or  current  surges  in  the  screen  circuit 
are  apt  to  permanently  damage  the  tube. 

The  Series  Screen  A simple  method  of  obtain- 
Supply  ing  screen  voltage  is  by 

means  of  a dropping  resis- 
tor from  the  high  voltage  plate  supply,  as  shown 
in  figure  33.  Since  the  current  drawn  by  the 
screen  is  a function  of  the  exciting  voltage 
applied  to  the  tetrode,  the  screen  voltage  will 
rise  to  equal  the  plate  voltage  under  condi- 
tions of  no  exciting  voltage.  If  the  control  grid 
is  overdriven,  on  the  other  hand,  the  screen 
current  may  become  excessive.  In  either  case, 
damage  to  the  screen  and  its  associated  com- 
ponents may  result.  In  addition,  fluctuations 
in  the  plate  loading  of  the  tetrode  stage  will 
cause  changes  in  the  screen  current  of  the 
tube.  This  will  result  in  screen  voltage  fluc- 
tuations due  to  the  inherently  poor  voltage 
regulation  of  the  screen  series  dropping  resis- 
tor. These  effects  become  dangerous  to  tube 
life  if  the  plate  voltage  is  greater  than  the 
screen  voltage  by  a factor  of  2 or  so. 


270  Generation  of  R-F  E 


nergy 


THE  RADIO 


The  Clamp  Tube  A clamp  tube  may  be  added 
to  the  series  screen  supply, 
as  shown  in  figure  34.  The  clamp  tube  is  nor- 
mally cut  off  by  virtue  of  the  d-c  grid  bias  drop 
developed  across  the  grid  resistor  of  the  tet- 
rode tube.  When  excitation  is  removed  from 
the  tetrode,  no  bias  appears  across  the  grid 
resistor,  and  the  clamp  tube  conducts  heavily, 
dropping  the  screen  voltage  to  a safe  value. 
When  excitation  is  applied  to  the  tetrode  the 
clamp  tube  is  inoperative,  and  fluctuations  of 
the  plate  loading  of  the  tetrode  tube  could 
allow  the  screen  voltage  to  rise  to  a damaging 
value.  Because  of  this  factor,  the  clamp  tube 
does  not  offer  complete  protection  to  the  tet- 
rode. 

The  Separate  A low  voltage  screen  supply 
Screen  Supply  may  be  used  instead  of  the 
series  screen  dropping  resis- 
tor. This  will  protect  the  screen  circuit  from 
excessive  voltages  when  the  other  tetrode 
operating  parameters  shift.  However,  the  screen 
can  be  easily  damaged  if  plate  or  bias  volt- 
age is  removed  from  the  tetrode,  as  the  screen 
current  will  reach  high  values  and  the  screen 
dissipation  will  be  exceeded.  If  the  screen 
supply  is  capable  of  providing  slightly  more 
screen  voltage  than  the  tetrode  requires  for 
proper  operation,  a series  wattage-limiting  re- 
sistor may  be  added  to  the  circuit  as  shown 
in  figure  35.  With  this  resistor  in  the  circuit 
it  is  possible  to  apply  excitation  to  the  tet- 
rode tube  with  screen  voltage  present  (but  in 
the  absence  of  plate  voltage)  and  still  not  dam- 
age the  screen  of  the  tube.  The  value  of  the 
resistor  should  be  chosen  so  that  the  product 
of  the  voltage  applied  to  the  screen  of  the 
tetrode  times  the  screen  current  never  exceeds 
the  maximum  rated  screen  dissipation  of  the 
tube. 


13-13  Interstage  Coupling 

Energy  is  usually  coupled  from  one  circuit 
of  a transmitter  into  another  either  by  capaci- 
tive coupling,  inductive  coupling,  or  link  cou- 


Figure 34 

CLAMP- TUBE  SCREEN  SUPPLY 


pling.  The  latter  is  a special  form  of  induc- 
tive coupling.  The  choice  of  a coupling  method 
depends  upon  the  purpose  for  which  it  is  to 
be  used. 

Capacitive  Capacitive  coupling  between  an 
Coupling  amplifier  or  doubler  circuit  and  a 
preceding  driver  stage  is  shown 
in  figure  36.  The  coupling  capacitor,  C,  iso- 
lates the  d-c  plate  supply  from  the  next  grid 
and  provides  a low  impedance  path  for  the  r-f 
energy  between  the  tube  being  driven  and  the 
driver  tube.  This  method  of  coupling  is  simple 
and  economical  for  low  powet  amplifier  or  ex- 
citer stages,  but  has  certain  disadvantages, 
particularly  for  high  frequency  stages.  The 
grid  leads  in  an  amplifier  should  be  as  short 
as  possible,  but  this  is  difficult  to  attain  in 
the  physical  arrangement  of  a high  power  am- 
plifier with  respect  to  a capacitively-coupled 
driver  stage. 

Disadvantages  of  One  significant  disadvan- 
Copocitive  tage  of  capacitive  coupling 

Coupling  is  the  difficulty  of  adjusting 

the  load  on  the  driver  stage. 
Impedance  adjustment  can  be  accomplished 
by  tapping  the  coupling  lead  a part  of  the  way 
down  on  the  plate  coil  of  the  tuned  stage  of 
the  driver  circuit;  but  often  when  this  is  done 


LOW  VOLTAGE  t-B 
SCREEN  SUPPLY 


Figure  35 

A PROTECTIVE  WATTAGE-LIMITING  RE- 
SISTOR FOR  USE  WITH  LOW-VOLTAGE 
SCREEN  SUPPLY 


HANDBOOK 


Interstage  Coupling  271 


Figure  36 

CAPACITIVE  INTERSTAGE  COUPLING 


a parasitic  oscillation  will  take  place  in  the 
stage  being  driven. 

One  main  disadvantage  of  capacitive  coupl- 
ing lies  in  the  fact  that  the  grid-to-filament 
capacitance  of  the  driven  tube  is  placed  di- 
rectly across  the  driver  tuned  circuit.  This 
condition  sometimes  makes  the  r-f  amplifier 
difficult  to  neutralize,  and  the  increased  mini- 
mum circuit  capacitance  makes  it  difficult  to 
use  a reasonable  size  coil  in  the  v-h-f  range. 
Difficulties  from  this  source  can  be  partially 
eliminated  by  using  a center-tapped  or  split- 
stator  tank  circuit  in  the  plate  of  the  driver 
stage,  and  coupling  capacitively  to  the  oppo- 
site end  from  the  plate.  This  method  places 
the  plate-to-filament  capacitance  of  the  driver 
across  one-half  of  the  tank  and  the  grid-to- 
filament  capacitance  of  the  following  stage 
across  the  other  half.  This  type  of  coupling  is 
shown  in  figure  37. 

Capacitive  coupling  can  be  used  to  advan- 
tage in  reducing  the  total  number  of  tuned  cir- 
cuits in  a transmitter  so  as  to  conserve  space 
and  cost.  It  also  can  be  used  to  advantage  be- 
tween stages  for  driving  beam  tetrode  or  pen- 
tode amplifier  or  doubler  stages. 

Inductive  Inductive  coupling  (figure  38)  re- 
Coupling  suits  when  two  coils  are  electro- 
magnetically  coupled  to  one  an- 
other. The  degree  of  coupling  is  controlled  by 
varying  the  mutual  inductance  of  the  two  coils, 
which  is  accomplished  by  changing  the  spac- 
ing or  the  relationship  between  the  axes  of 
the  coils. 


Figure  38 

INDUCTIVE  INTERSTAGE  COUPLING 


BALANCED  CAPACITIVE  COUPLING 


Balanced  capacitive  coupling  sometimes  is 
useful  when  it  is  desirable  to  use  a rel atively 
large  inductance  in  the  interstage  tank  cir- 
cuit, or  where  the  exciting  stage  is  neutral- 
ized as  shown  above. 


Inductive  coupling  is  used  extensively  for 
coupling  r-f  amplifiers  in  radio  receivers.  How- 
ever, the  mechanical  problems  involved  in  ad- 
justing the  degree  of  coupling  limit  the  use- 
fulness of  direct  inductive  coupling  in  trans- 
mitters. Either  the  primary  or  the  secondary 
or  both  coils  may  be  tuned. 

Unity  Coupling  If  the  grid  tuning  capacitor  of 
figure  38  is  removed  and  the 
coupling  increased  to  the  maximum  practicable 
value  byinterwindingthe  turns  of  the  two  coils, 
the  circuit  insofar  as  r.f.  is  concerned  acts 
like  that  of  figure  36,  in  which  one  tank  serves 
both  as  plate  tank  for  the  driver  and  grid  tank 
for  the  driven  stage.  The  inter-wound  grid 
winding  serves  simply  to  isolate  the  d-c  plate 
voltage  of  the  driver  from  the  grid  of  the  driven 
stage,  and  to  provide  a return  for  d-c  grid  cur- 
rent. This  type  of  coupling,  illustrated  in  fig- 
ure 39,  is  commonly  known  as  unity  coupling. 

Because  of  the  high  mutual  inductance,  both 
primary  and  secondary  are  resonated  by  the 
one  tuning  capacitor. 


INTERWOUND 


Figure  39 

"UNITY"  INDUCTIVE  COUPLING 

Due  to  the  high  value  of  coppl/ng  between 
the  two  collSf  one  tuning  capacitor  tunes 
both  circuits.  This  arrangement  often  is  use- 
ful In  coupling  from  a single-ended  to  a push- 
pull  stage. 


272  Generation  of  R-F  Energy 


THE  RADIO 


Figure  40 

INTERSTAGE  COUPLING  BY  MEANS 
OF  A "LINK" 

Link  interstage  coupling  is  very  commonly 
used  since  the  two  stages  may  be  separated 
by  a considerable  distance,  since  the  amount 
of  a coupling  between  the  two  stages  may  be 
easily  varied,  and  since  fhe  capacitances  of 
the  two  stages  may  be  isolated  to  permit  use 
of  larger  in</oc#ances  in  the  v~h-f  ronge. 


Link  Coupling  A special  form  of  inductive 
coupling  which  is  widely  em- 
ployed in  radio  transmitter  circuits  is  known 
as  link  coupling.  A low  impedance  r-f  trans- 
mission line  couples  the  two  tuned  circuits 
together.  Each  end  of  the  line  is  terminated 
in  one  or  more  turns  of  wire,  or  links,  wound 
around  the  coils  which  are  being  coupled  to- 
gether. These  links  should  be  coupled  to  each 
tuned  circuit  at  the  point  of  zero  r-f  potential, 
or  nodal  point.  A ground  connection  to  one 
side  of  the  link  usually  is  used  to  reduce  har- 
monic coupling,  or  where  capacitive  coupling 
between  two  circuits  must  be  minimized.  Co- 
axial line  is  commonly  used  to  transfer  energy 
between  the  two  coupling  links,  although  Twin- 
Lead  may  be  used  where  harmonic  attenuation 
is  not  so  important. 

Typical  link  coupled  circuits  are  shown  in 
figures  40  and  41.  Some  of  the  advantages  of 
link  coupling  are  the  following: 

(1)  It  eliminates  coupling  taps  on  tuned  cir- 
cuits. 

(2)  It  permits  the  use  of  series  power  supply 
connections  in  both  tuned  grid  and  tuned 
plate  circuits,  and  thereby  eliminates  the 
need  of  shunt-feed  r-f  chokes. 

(3)  It  allows  considerable  separation  between 
transmitter  stages  without  appreciable 
r-f  losses  or  stray  chassis  currents. 

(4)  It  reduces  capacitive  coupling  and  there- 
by makes  neutralization  more  easily  at- 
tainable in  r-f  amplifiers. 

(5)  It  provides  semi-automatic  impedance 
matching  between  plate  and  grid  tuned 
circuits,  with  the  result  that  greater  grid 
drive  can  be  obtained  in  comparison  to 
capacitive  coupling. 

(6)  It  effectively  reduces  the  coupling  of  har- 
monic energy. 


Figure  41 

PUSH-PULL  LINK  COUPLING 


The  link-coupling  line  and  links  can  be 
made  of  no.  18  push-back  wire  for  coupling 
between  low-power  stages.  For  coupling  be- 
tween higher  powered  stages  the  150-ohm 
Twin-Lead  transmission  line  is  quite  effective 
and  has  very  low  loss.  Coaxial  transmission  is 
most  satisfactory  between  high  powered  am- 
plifier stages,  and  should  always  be  used 
where  harmonic  attenuation  is  important. 


13-14  Rodio-Frequency 
Chokes 

Radio-frequency  chokes  are  connected  in 
circuits  for  the  purpose  of  stopping  the  pas- 
sage of  r-f  energy  while  still  permitting  a di- 
rect current  or  audio-frequency  current  to  pass. 
They  consist  of  inductances  wound  with  a 
large  number  of  turns,  either  in  the  form  of  a 
solenoid,  a series  of  solenoids,  a single  uni- 
versal pie  winding,  or  a series  of  pie  wind- 
ings. These  inductors  are  designed  to  have  as 
much  inductance  and  as  little  distributed  or 
shunt  capacitance  as  possible.  The  unavoid- 
able small  amount  of  distributed  capacitance 
resonates  the  inductance,  and  this  frequency 
normally  should  be  much  lower  than  the  fre- 
quency at  which  the  transmitter  or  receiver 
circuit  is  operating.  R-f  chokes  for  operation 
on  several  bands  must  be  designed  carefully 
so  that  the  impedance  of  the  choke  will  be  ex- 
tremely high  (several  hundred  thousand  ohms) 
in  each  of  the  bands. 

The  direct  current  which  flows  through  the 
r-f  choke  largely  determines  the  size  of  wire 
to  be  used  in  the  winding.  The  inductance  of 
r-f  chokes  for  the  v-h-f  range  is  much  less 
than  for  chokes  designed  for  broadcast  and 
ordinary  short-wave  operation,  A very  high 
inductance  r-f  choke  has  more  distributed  ca- 
pacitance than  a smaller  one,  with  the  result 


HANDBOOK 


Shunt  and  Series  Feed  273 


PARALLEL  PLATE  FEED  SERIES  PLATE  FEED 


Figure  42 

ILLUSTRATING  PARALLEL  AND 
SERIES  PLATE  FEED 

Parallel  plate  feed  is  desirable  from  a safety 
standpoint  since  the  tank  circuit  is  at  ground 
potential  with  respect  to  d.c.  However,  a 
high-impedance  r-f  choke  is  required,  and 
the  r-f  choke  must  be  able  to  withstand  the 
peak  r-f  voltage  output  of  the  tube.  Series 
plate  feed  eliminates  the  requirement  for  a 
high-performance  r-f  choke,  but  requires  the 
use  of  a relatively  large  value  of  by-pass 
capacitance  at  the  bottom  end  of  the  tank 
circuit,  as  contrasted  to  the  moderate  value 
of  coupling  capacitance  which  may  be  used 
at  the  top  of  the  tank  circuit  for  parallel 
plate  feed. 


thac  it  will  actually  offer  less  impedance  at 
very  high  frequencies. 

Another  consideration,  just  as  important  as 
the  amount  of  d.c,  the  winding  will  carry,  is 
the  r-f  voltage  which  may  be  placed  across 
the  choke  without  its  breaking  down.  This  is 
a function  of  insulation,  turn  spacing,  frequen- 
cy, number  and  spacing  of  pies  and  other  fac- 
tors. 

Some  chokes  which  are  designed  to  have  a 
high  impedance  over  a very  wide  range  of  fre- 
quency are,  in  effect,  really  two  chokes:  a 
u-h-f  choke  in  series  with  a high-frequency 
choke.  A choke  of  this  type  is  polarized;  that 
is,  it  is  important  that  the  correct  end  of  the 
combination  choke  be  connected  to  the  **hot” 
side  of  the  circuit. 

Shunt  and  Direct-current  grid  and  plate 

Series  Feed  connections  are  made  either  by 
series  or  parallel  feed  systems. 
Simplified  forms  of  each  are  shown  in  figures 

42  and  43. 

Series  feed  can  be  defined  as  that  in  which 
the  d-c  connection  is  made  to  the  grid  or  plate 
circuits  at  a point  of  very  low  r-f  potential. 
Shunt  feed  always  is  made  to  a point  of  high 
r-f  voltage  and  always  requires  a high  imped- 
ance r-f  choke  or  a relatively  high  resistance 
to  prevent  waste  of  r-f  power. 


-BIAS 


PARALLEL  BIAS  FEED 


SERIES  BIAS  FEED 


Figure  43 

ILLUSTRATING  SERIES  AND 
PARALLEL  BIAS  FEED 


13-15  Parallel  and 

Push-Pull  Tube  Circuits 


The  comparative  r-f  power  output  from  paral- 
lel or  push-pulloperated  amplifiers  is  the  same 
if  proper  impedance  matching  is  accomplished, 
if  sufficient  grid  excitation  is  available  in 
both  cases,  and  if  the  frequency  of  measure- 
ment is  considerably  lower  than  the  frequency 
limit  of  the  tubes. 

Parallel  Operating  tubes  in  parallel  has 

Operation  some  advantages  in  transmitters 

designed  for  operation  below  10 
Me.,  particularly  when  tetrode  or  pentode  tubes 
are  to  be  used.  Only  one  neutralizing  capacitor 
is  required  for  parallel  operation  of  triode 
tubes,  as  against  two  for  push-pull.  Above 
about  10  Me.,  depending  upon  the  tube  type, 
parallel  tube  operation  is  not  ordinarily  recom- 
mended with  triode  tubes.  However,  parallel 
operation  of  grounded-grid  stages  and  stages 
using  low-C  beam  tetrodes  often  will  give  ex- 
cellent results  well  into  the  v-h-f  range. 

Push-Pull  The  push-pull  connection  provides 

Operation  a well-balanced  circuit  insofar  as 

miscellaneous  capacitances  are 
concerned;  in  addition,  the  circuit  can  be  neu- 
tralized more  completely,  especially  in  high- 
frequency  amplifiers.  The  L/C  ratio  in  a push- 
pull  amplifier  can  be  made  higher  than  in  a 
plate-neutralized  parallel-tube  operated  am- 
plifier. Push-pull  amplifiers,  when  perfectly 
balanced,  have  less  second-harmonic  output 
than  parallel  or  single-tube  amplifiers,  but  in 
practice  undesired  capacitive  coupling  and 
circuit  unbalance  more  or  less  offset  the  theo- 
retical harmonic-reducing  advantages  of  push- 
pull  r-f  circuits. 


CHAPTER  FOURTEEN 


R-F  Feedback 


Comparatively  high  gain  is  required  in  sin- 
gle sideband  equipment  because  the  signal  is 
usually  generated  at  levels  of  one  watt  or  less. 
To  get  from  this  level  to  a kilowatt  requires 
about  30  db  of  gain.  High  gain  tetrodes  may 
be  used  to  obtain  this  increase  with  a minimum 
number  of  stages  and  circuits.  Each  stage  con- 
tributes some  distortion;  therefore,  it  is  good 
practice  to  keep  the  number  of  stages  to  a 
minimum.  It  is  generally  considered  good  prac- 
tice to  operate  the  low  level  amplifiers  below 
their  maximum  power  capability  in  order  to 
confine  most  of  the  distortion  to  the  last  two 
amplifier  stages.  R-f  feedback  can  then  be 
utilized  to  reduce  the  distortion  in  the  last 
two  stages.  This  type!  of  feedback  is  no  dif- 
ferent from  the  common  audio  feedback  used 
in  high  fidelity  sound  systems.  A sample  of 
the  output  waveform  is  applied  to  the  ampli- 
fier input  to  correct  the  distortion  developed 
in  the  amplifier.  The  same  advantages  can  be 
obtained  at  radio  frequencies  that  are  obtained 
at  audio  frequencies  when  feedback  is  used. 


14-1  R-F  Feedback 
Circuits 

R-f  feedback  circuits  have  been  developed 
by  the  Collins  Radio  Co.  for  use  with  linear 
amplifiers.  Tests  with  large  receiving  and  small 
transmitting  tubes  showed  that  amplifiers  us- 
ing these  tubes  without  feedback  developed 
signal-to-distortion  ratios  no  better  than  30  db 
or  so.  Tests  were  run  employing  cathode  fol- 
lower circuits,  such  as  shown  in  figure  lA. 
Lower  distortion  was  achieved,  but  at  the  cost 
of  low  gain  per  stage.  Since  the  voltage  gain 
through  the  tube  is  less  than  unity,  all  gain 
has  to  be  achieved  by  voltage  step-up  in  the 
tank  circuits.  This  gain  is  limited  by  the  dis- 
sipation of  the  tank  coils,  since  the  circuit 
capacitance  across  the  coils  in  a typical  trans- 
mitter is  quite  high.  In  addition,  the  tuning 
of  such  a stage  is  sharp  because  of  the  high 
Q circuits. 

The  cathode  follower  performance  of  the 
tube  can  be  retained  by  moving  the  r-f  ground 


B-t- 


Figure  1 

SIMILAR  CATHODE  FOLLOWER  CIRCUITS  HAVING  DIFFERENT  R-F  GROUND  POINTS. 

274 


R-F  Feedback  Circuits  275 


SINGLE  STAGE  AMPLIFIER  WITH 
R-F  FEEDBACK  CIRCUIT 


Figure  3 

SINGLE  STAGE  FEEDBACK 
AMPLIFIER  WITH  GROUND 
RETURN  POINT  MODIFIED  FOR 
UNBALANCED  INPUT  AND 
OUTPUT  CONNECTIONS. 


point  of  the  circuit  from  the  plate  to  the  cath- 
ode as  shown  in  figure  IB.  Both  ends  of  the 
input  circuit  are  at  high  r-f  potential  so  in- 
ductive coupling  to  this  type  of  amplifier  is 
necessary. 

Inspection  of  figure  IB  shows  that  by  mov- 
ing the  top  end  of  the  input  tank  down  on  a 
voltage  divider  tap  across  the  plate  tank  cir- 
cuit, the  feedback  can  be  reduced  from  100%, 
as  in  the  case  of  the  cathode  follower  circuit, 
down  to  any  desired  value.  A typical  feedback 
circuit  is  illustrated  in  figure  2.  This  circuit 
is  more  practical  than  those  of  figure  1,  since 
the  losses  in  the  input  tank  are  greatly  reduced. 
A feedback  level  of  12  db  may  be  achieved 
as  a good  compromise  between  distortion  and 
stage  gain.  The  voltage  developed  across  & 
will  be  three  times  the  grid-cathode  voltage. 


B-l- 


RFcj 

! 

11  . f - 

k 

1 

1 

!ci5  j 

^ ^ V 
RFC3 

-L.  < 

N 

/ 

BIAS  B+ 


Figure  4 

R-F  AMPLIFIER  WITH  FEEDBACK 
AND  IMPEDANCE  MATCHING 
OUTPUT  NETWORK. 

Tuning  and  loading  are  accomplished  by  Ct 
and  C.U  Ci  and  Li  are  tuned  in  unison  to 
establish  the  correct  degree  of  feedback. 


Inductive  coupling  is  required  for  this  cir- 
cuit, as  shown  in  the  illustration. 

The  circuit  of  figure  3 eliminates  the  need 
for  inductive  coupling  by  moving  the  r-f 
ground  to  the  point  common  to  both  tank 
circuits.  The  advantages  of  direct  coupling  be- 
tween stages  far  outweigh  the  disadvantages  of 
having  the  r-f  feedback  voltage  appear  on  the 
cathode  of  the  amplifier  tube. 

In  order  to  match  the  amplifier  to  a load, 
the  circuit  of  figure  4 may  be  used.  The  ratio 
of  XLi  to  XCi  determines  the  degree  of  feed- 
back, so  it  is  necessary  to  tune  them  in  unison 
when  the  frequency  of  operation  is  changed. 
Tuning  and  loading  functions  are  accomplished 
by  varying  Ci  and  Cs.  L2  may  also  be  varied  to 
adjust  the  loading. 

Feedback  Around  a The  maximum  phase 

Two-Stage  Amplifier  shift  obtainable  over 

two  simple  tuned  cir- 
cuits does  not  exceed  180  degrees,  and  feed- 
back around  a two  stage  amplifier  is  possible. 
The  basic  circuit  of  a two  stage  feedback 
amplifier  is  shown  in  figure  5.  This  circuit 
is  a conventional  two-stage  tetrode  amplifier 
except  that  r-f  is  fed  back  from  the  plate 
circuit  of  the  PA  tube  to  the  cathode  of  the 
driver  tube.  This  will  reduce  the  distortion 


Figure  5 

BASIC  CIRCUIT  OF  TWO-STAGE  AMPLIFIER  WITH  R-F  FEEDBACK 


Feedback  voltage  is  obtained  from  a voltage  divider  across  the  output  circuit  end 
applied  directly  to  the  cathode  of  the  first  tube.  The  input  tank  circuit  is  thus 
outside  the  feedback  loop. 


276  R-F  Feedback 


THE  RADIO 


of  both  tubes  as  effectively  as  using  individual 
feedback  loops  around  each  stage,  yet  will 
allow  a higher  level  of  overall  gain.  With 
only  two  tuned  circuits  in  the  feedback  loop, 
it  is  possible  to  use  12  to  15  db  of  feedback 
and  still  leave  a wide  margin  for  stability.  It 
is  possible  to  reduce  the  distortion  by  nearly 
as  many  db  as  are  used  in  feedback.  This  cir- 
cuit has  two  advantages  that  are  lacking  in  the 
single  stage  feedback  amplifier.  First,  the  fila- 
ment of  the  output  stage  can  now  be  operated 
at  r-f  ground  potential.  Second,  any  conven- 
tional pi  output  network  may  be  used. 

R-f  feedback  will  correct  several  types  of 
distortion.  It  will  help  correct  distortion  caused 
by  poor  power  supply  regulation,  too  low  grid 
bias,  and  limiting  on  peaks  when  the  plate 
voltage  swing  becomes  too  high. 

Neutralization  The  purpose  of  neutraliza- 

and  R-F  Feedback  tion  of  an  r-f  amplifier 
Stage  is  to  balance  out  ef- 
fects of  the  grid-plate  capacitance  coupling  in 
the  amplifier.  In  a conventional  amplifier  us- 
ing a tetrode  tube,  the  effective  input  capacity 
is  given  by: 

Input  Capacitance  = Cm  + Cgp  ( 1 + A cos  *) 

where:  Cm  = tube  input  capacitance 
Csp  — grid-plate  capacitance 
A = voltage  amplification  from  grid 
to  plate 

S = phase  angle  of  load 

In  a typical  unneutralized  tetrode  amplifier 
having  a stage  gain  of  33)  the  input  capaci- 
tance of  the  tube  with  the  plate  circuit  in 
resonance  is  increased  by  8 ^i^fd.  due  to  the 
unneutralized  grid-plate  capacitance.  This  is 
unimportant  in  amplifiers  where  the  gain  (A) 
remains  constant  but  if  the  tube  gain  varies, 
serious  detuning  and  r-f  phase  shift  may  result. 
A grid  or  screen  modulated  r-f  amplifier  is  an 
example  of  the  case  where  the  stage  gain  var- 
ies from  a maximum  down  to  zero.  The  gain 
of  a tetrode  r-f  amplifier  operating  below  plate 
current  saturation  varies  with  loading  so  that 
if  it  drives  a following  stage  into  grid  current 
the  loading  increases  and  the  gain  falls  off. 

The  input  of  the  grid  circuit  is  also  affected 
by  the  grid-plate  capacitance,  as  shown  in  this 
equation: 

Input  Resistance  = ^ ^ - 

2i7f  X Cbp  (Asin«) 

This  resistance  is  in  shunt  with  the  grid 
current  loading,  grid  tank  circuit  losses,  and 
driving  source  impedance.  When  the  plate  cir- 


cuit is  inductive  there  is  energy  transferred 

from  the  plate  to  the  grid  circuit  (positive 
feedback)  which  will  introduce  negative  resist- 
ance in  the  grid  circuit.  When  this  shunt 
negative  resistance  across  the  grid  circuit  is 
lower  than  the  equivalent  positive  resistance 
of  the  grid  loading,  circuit  losses,  and  driving 
source  impedance,  the  amplifier  will  oscillate. 

When  the  plate  circuit  is  in  resonance 
(phase  angle  equal  to  zero)  the  input  resist- 
ance due  to  the  grid-plate  capacitance  becomes 
infinite.  As  the  plate  circuit  is  tuned  to  the 
capacitive  side  of  resonance,  the  input  resist- 
ance becomes  positive  and  power  is  actually 
transferred  from  the  grid  to  the  plate  circuit. 
This  is  the  reason  that  the  grid  current  in  an 
unneutralized  tetrode  r-f  amplifier  varies  from 
a low  value  with  the  plate  circuit  tuned  on  the 
low  frequency  side  of  resonance  to  a high  value 
on  the  high  frequency  side  of  resonance  The 
grid  current  is  proportional  to  the  r-f  voltage 
on  the  grid  which  is  varying  under  these  con- 
ditions. In  a tetrode  class  ABi  amplifier,  the 
effect  of  grid-plate  feedback  can  be  observed 
by  placing  a r-f  voltmeter  across  the  grid  cir- 
cuit and  observing  the  voltage  change  as  the 
plate  circuit  is  tuned  through  resonance. 

If  the  amplifier  is  over-neutralized,  the  ef- 
fects reverse  so  that  with  the  plate  circuit 
tuned  to  the  low  frequency  side  of  resonance 
the  grid  voltage  is  high,  and  on  the  high  fre- 
quency side  of  resonance,  it  is  low. 

Amplifier  A useful  "rule  of 

Neutralization  Check  thumb"  method  of 

checking  neutraliza- 
tion of  an  amplifier  stage  (assuming  that  it 
is  nearly  correct  to  start  with)  is  to  tune  both 
grid  and  plate  circuits  to  resonance.  Then,  ob- 
serving the  r-f  grid  current,  tune  the  plate  cir- 
cuit to  the  high  frequency  side  of  resonance. 
If  the  grid  current  rises,  more  neutralization 
capacitance  is  required.  Conversely,  if  the  grid 
current  decreases,  less  capacitance  is  needed. 
This  indication  is  very  sensitive  in  a neutral- 
ized triode  amplifier,  and  correct  neutraliza- 
tion exists  when  the  grid  current  peaks  at  the 
point  of  plate  current  dip.  In  tetrode  power 
amplifiers  this  indication  is  less  pronounced. 
Sometimes  in  a supposedly  neutralized  tetrode 
amplifier,  there  is  practically  no  change  in 
grid  voltage  as  the  plate  circuit  is  tuned 
through  resonance,  and  in  some  amplifiers  it 
is  unchanged  on  one  side  of  resonance  and 
drops  slightly  on  the  other  side.  Another  ob- 
servation sometimes  made  is  a small  dip  in 
the  center  of  a broad  peak  of  grid  current. 
These  various  effects  are  probably  caused  by 


HANDBOOK 


R-F  Feedback  Circuits  277 


Figure  6 

SINGLE  STAGE  R-F  AMPLIFIER 
WITH  FEEDBACK  RATIO  OF 


C./C,  to  C„p/C,/  DETERMINES 
STAGE  NEUTRALIZATION 


Figure  7 

NEUTRALIZED  AMPLIFIER  AND 
INHERENT  FEEDBACK  CIRCUIT. 


Neutralization  is  achieved  by  varying 
the  capacity  of  Cn, 


coupling  from  the  plate  to  the  grid  circuit 
through  other  paths  which  are  not  balanced 
out  by  the  particular  neutrali2ing  circuit  used. 

Feedback  and  Figure  6 shows  an  r-f  am- 
Neutralization  plifier  with  negative  feed- 
af  a One-Stage  back.  The  voltage  developed 
R-F  Amplifier  across  C«  due  to  the  voltage 
divider  action  of  Ca  and  Ci 
is  introduced  in  series  with  the  voltage  devel- 
oped across  the  grid  tank  circuit  and  is  in 
phase-opposition  to  it.  The  feedback  can  be 
made  any  value  from  zero  to  100%  by  proper- 
ly choosing  the  values  of  Ca  and  Ca. 

For  reasons  stated  previously,  it  is  necessary 
to  neutralize  this  amplifier,  and  the  relation- 
ship for  neutralization  is: 

Ca  Cgp 

Ca  Cpf 

It  is  often  necessary  to  add  capacitance  from 
plate  to  grid  to  satisfy  this  relationship 

Figure  7 is  identical  to  figure  6 except  that 
it  is  redrawn  to  show  the  feedback  inherent  in 
this  neutralization  circuit  more  clearly.  Cn  and 
C replace  Ca  and  Ci,  and  the  main  plate  tank 
tuning  capacitance  is  Cs.  The  circuit  of  figure 
7 presents  a problem  in  coupling  to  the  grid 
circuit.  Inductive  coupling  is  ideal,  but  the 
extra  tank  circuits  complicate  the  mning  of  a 
transmitter  which  uses  several  cascaded  am- 
plifiers with  feedback  around  each  one.  The 
grid  could  be  coupled  to  a high  source  imped- 
ance such  as  a tetrode  plate,  but  the  driver 
then  cannot  use  feedback  because  this  would 
cause  the  source  impedance  to  be  low.  A pos- 
sible solution  is  to  move  the  circuit  ground 
point  from  the  cathode  to  the  bottom  end  of 
the  grid  tank  circuit.  The  feedback  voltage  then 
appears  between  the  cathode  and  ground 
( figure  8 ) . The  input  can  be  capacitively 
coupled,  and  the  plate  of  the  amplifier  can 
be  capacitively  coupled  to  the  next  stage.  Also, 
cathode  type  transmitting  tubes  are  available 
that  allow  the  heater  to  remain  at  ground  po- 


tential when  r-f  is  impressed  upon  the  cathode. 
The  output  voltage  available  with  capacity 
coupling,  of  course,  is  less  than  the  plate- 
cathode  r-f  voltage  developed  by  the  amount 
of  feedback  voltage  across  Ci. 

14-2  Feedback  and 

Neutralizafion  of  a 
Two-Sfage  R-F  Amplifier 

Feedback  around  two  r-f  stages  has  the  ad- 
vantage that  more  of  the  tube  gain  can  be 
realized  and  nearly  as  much  distortion  reduc- 
tion can  be  obtained  using  12  db  around  two 
stages  as  is  realized  using  12  db  around  each 
of  two  stages  separately.  Figure  9 shows  a 
basic  circuit  of  a two  stage  feedback  ampli- 
fier, Inductive  output  coupling  is  used,  al- 
though a pi-network  configuration  will  also 
work  well.  The  small  feedback  voltage  required 
is  obtained  from  the  voltage  divider  Ci  - & 
and  is  applied  to  the  cathode  of  the  driver 
tube.  Cl  is  only  a few  ggfd.,  so  this  feedback 
voltage  divider  may  be  left  fixed  for  a wide 
frequency  range.  If  the  combined  tube  gain  is 
160,  and  12  db  of  feedback  is  desired,  the  ratio 
of  C2  to  Ci  is  about  40  to  1.  This  ratio  in 
practice  may  be  400  ggfd.  to  2.5  ggfd.,  for 
example. 

A complication  is  introduced  into  this  sim- 
plified circuit  by  the  cathode-grid  capacitance 


Figure  8 

UNBALANCED  INPUT  AND  OUTPUT 
CIRCUITS  FOR  SINGLE-STAGE 
R-F  AMPLIFIER  WITH  FEEDBACK 


278  R-F  Feedback 


THE  RADIO 


V1 


V2 


Figure  9 

TWO-STAGE  AMPLIFIER  WITH  FEEDBACK. 

Included  is  a capacitor  (Cs)  lor  neutralizing  the  cathode-grid  capacity  at  the  first  tube.  Vi  is  neutralized 
by  capacitor  Cs,  and  Vr  is  neutralized  by  the  correct  ratio  of  C,/Ct. 


of  the  first  tube  which  causes  an  undersired 
coupling  to  the  input  grid  circuit.  It  is  neces- 
sary to  neutralize  out  this  capacitance  coupling, 
as  illustrated  in  figure  9-  The  relationship  for 
neutralization  is: 

Cs  Cgt 

Cs  Cs 

The  input  circuit  may  be  made  unbalanced 
by  making  Cs  five  times  the  capacity  of  Cs. 
This  will  tend  to  reduce  the  voltage  across 
the  coil  and  to  minimize  the  power  dissipated 
by  the  coil.  For  proper  balance  in  this  case, 
C»  must  be  five  times  the  grid-filament  capaci- 
tance of  the  tube. 

Except  for  tubes  having  extremely  small 
grid-plate  capacitance,  it  is  still  necessary  to 
properly  neutralize  both  tubes.  If  the  ratio  of 
Cl  to  Cs  is  chosen  to  be  equal  to  the  ratio  of 
the  grid-plate  capacitance  to  the  grid- filament 
capacitance  in  the  second  tube  (Vs),  this  tube 
will  be  neutralized.  Tubes  such  as  a 4X-150A 
have  very  low  grid-plate  capacitance  and  prob- 
ably will  not  need  to  be  neutralized  when  used 
in  the  first  (Vi)  stage.  If  neutralization  is 
necessary,  capacitor  Cs  is  added  for  this  pur- 
pose and  the  proper  value  is  given  by  the 
following  relationship: 

Cgp  Cgf  Cs 

Cs  Cs  Cl 

If  neither  tube  requires  neutralization,  the 
bottom  end  of  the  interstage  tank  circuit  may 
be  returned  to  r-f  ground.  The  screen  and 
suppressor  of  the  first  tube  should  then  be 
grounded  to  keep  the  tank  output  capaci- 
tance directly  across  this  interstage  circuit  and 
to  avoid  common  coupling  between  the  feed- 
back on  the  cathode  and  the  interstage  circuit. 
A slight  amount  of  degeneration  occurs  in  the 
first  stage  since  the  tube  also  acts  as  a grounded 
grid  amplifier  with  the  screen  as  the  grounded 
grid.  The  /r  of  the  screen  is  much  lower  than 
that  of  the  control  grid  so  that  this  effect  may 
be  unnoticed  and  would  only  require  slightly 


more  feedback  from  the  output  stage  to  over- 
come. 

Tests  For  Neutralizing  the  circuit  of 

Neutralization  figure  9 balances  out  cou- 
pling between  the  input 
tank  circuit  and  the  output  tank  circuit,  but  it 
does  not  remove  all  coupling  from  the  plate 
circuit  to  the  grid-cathode  tube  input.  This 
latter  coupling  is  degenerative,  so  applying  a 
signal  to  the  plate  circuit  will  cause  a signal 
to  appear  between  grid  and  cathode,  even 
though  the  stage  is  neutralized.  A bench  test 
for  neutralization  is  to  apply  a signal  to  the 
plate  of  the  tube  and  detect  the  presence  of  a 
signal  in  the  grid  coil  by  inductive  coupling 
to  it.  No  signal  will  be  present  when  the  stage 
is  neutralized.  Of  course,  a signal  could  be  in- 
ductively coupled  to  the  input  and  neutraliza- 
tion accomplished  by  adjusting  one  branch  of 
the  neutralizing  circuit  bridge  (Cs  for  ex- 
ample) for  minimum  signal  on  the  plate  cir- 
cuit. 

Neutralizing  the  cathode-grid  capacitance  of 
the  first  stage  of  figure  9 may  be  accomplished 
by  applying  a signal  to  the  cathode  of  the  tube 
and  adjusting  the  bridge  balance  for  minimum 
signal  on  a detector  inductively  coupled  to  the 
input  coil. 

Tuning  a Two-Stage  Tuning  the  two-stage 

Feedback  Amplifier  feedback  amplifier  of 

figure  9 is  accom- 
plished in  an  unconventional  way  because  the 
output  circuit  cannot  be  tuned  for  maximum 
output  signal.  This  is  because  the  output  cir- 
cuit must  be  tuned  so  the  feedback  voltage 
applied  to  the  cathode  is  in-phase  with  the 
input  signal  applied  to  the  first  grid.  When 
the  feedback  voltage  is  not  in-phase,  the  result- 
ant grid-cathode  voltage  increases  as  shown 
in  figure  10.  When  the  output  circuit  is 
properly  tuned,  the  resultant  grid-cathode  volt- 
age on  the  first  tube  will  be  at  a minimum,  and 
the  voltage  on  the  interstage  tuned  circuit  will 
also  be  at  a minimum. 


HANDBOOK 


Neutralization  279 


VOLTAGE- 
INPUT  GRID 
TO 

GROUND 


Figure  12 

INTERSTAGE  CIRCUIT  WITH 
SEPARATE  NEUTRALIZING 
AND  FEEDBACK  CIRCUITS. 


Figure  10 

VECTOR  RELATIONSHIP  OF 
FEEDBACK  VOLTAGE 

A~Ouiput  Circuit  Properly  Tuned 
B — Output  Circuit  Mis-Tuned 


The  two-stage  amplifier  may  be  tuned  by 
placing  a r-f  voltmeter  across  the  interstage 
tank  circuit  ("hot”  side  to  ground)  and  tuning 
the  input  and  interstage  circuits  for  maximum 
meter  reading,  and  tuning  the  output  circuit 
for  minimum  meter  reading.  If  the  second  tube 
is  driven  into  the  grid  current  region,  the  grid 
current  meter  may  be  used  in  place  of  the  r-f 
voltmeter.  On  high  powered  stages  where  oper- 
ation is  well  into  the  Class  AB  region,  the 
plate  current  dip  of  the  output  tube  indicates 
correct  output  circuit  tuning,  as  in  the  usual 
amplifier. 

Parasitic  Oscillations  in  Quite  often  low  fre- 
the  Feedback  Amplifier  q u e n c y parasitics 
may  be  found  in 
the  interstage  circuit  of  the  two-stage  feedback 
amplifier.  Oscillation  occurs  in  the  first  stage 
due  to  low  frequency  feedback  in  the  cathode 
circuit.  R-f  chokes,  coupling  capacitors,  and 
bypass  capacitors  provide  the  low  frequency 
tank  circuits.  When  the  feedback  and  second 
stage  neutralizing  circuits  are  combined,  it  is 
necessary  to  use  the  configuration  of  figure  11. 
This  circuit  has  the  advantage  that  only  one 
capacitor  (G)  is  required  from  the  plate  of 
the  output  tube,  thus  keeping  the  added  ca- 
pacitance across  the  output  tank  at  a minimum. 


Figure  1 1 

INTERSTAGE  CIRCUIT  COMBINING 
NEUTRALIZATION  AND 
FEEDBACK  NETWORKS. 


It  is  convenient,  however,  to  separate  these  cir- 
cuits so  neutralization  and  feedback  can  be 
adjusted  independently.  Also,  it  may  be  de- 
sirable to  be  able  to  switch  the  feedback  out 
of  the  circuit.  For  these  reasons,  the  circuit 
shown  in  figure  12  is  often  used.  Switch  Si 
removes  the  feedback  loop  when  it  is  closed. 

A slight  tendency  for  low  frequency  para- 
sitic oscillations  still  exists  with  this  circuit. 
Li  should  have  as  little  inductance  as  possible 
without  upsetting  the  feedback.  If  the  value  of 
Li  is  too  low,  it  cancels  out  part  of  the  re- 
actance of  feedback  capacitor  Ci  and  causes 
the  feedback  to  increase  at  low  values  of  radio 
frequency.  In  some  cases,  a swamping  resistor 
may  be  necessary  across  Li.  The  value  of  this 
resistor  should  be  high  compared  to  the  re- 
actance of  Cl  to  avoid  phase-shift  of  the  r-f 
feedback. 

14-3  Neutralization 
Procedure  in 

Feedback-Type  Amplifiers 

Experience  with  feedback  amplifiers  has 
brought  out  several  different  methods  of  neu- 
tralizing. An  important  observation  is  that 
when  all  three  neutralizing  adjustments  are 
correctly  made  the  peaks  and  dips  of  various 
tuning  meters  all  coincide  at  the  point  of  cir- 
cuit resonance.  For  example,  the  coincident  in- 
dications when  the  various  tank  circuits  are 
tuned  through  resonance  with  feedback  oper- 
ating are: 

A — When  the  PA  plate  circuit  is  tuned 
through  resonance: 

1 —  PA  plate  current  dip 

2 —  Power  output  peak 

3 —  PA  r-f  grid  voltage  dip 

4 —  PA  grid  current  dip 

(Note;  The  PA  grid  current  peaks 
when  feedback  circuit  is  disabled 
and  the  tube  is  heavily  driven) 


280  R-F  Feedback 


THE  RADIO 


Cn 


B — When  the  PA  grid  circuit  is  tuned 
through  resonance: 

1 —  Driver  plate  current  dip 

2 —  PA  r-f  grid  voltage  peak 

3 —  PA  grid  current  peak 

4 —  PA  power  output  peak 

C — When  the  driver  grid  circuit  is  tuned 
through  resonance: 

1 —  Driver  r-f  grid  voltage  peak 

2 —  Driver  plate  current  peak 

3 —  PA  r-f  grid  current  peak 

4 —  PA  plate  current  peak 

5 —  PA  power  output  peak 

Four  meters  may  be  employed  to  measure 
the  most  important  of  these  parameters.  The 
meters  should  be  arranged  so  that  the  follow- 
ing pairs  of  readings  are  displayed  on  meters 
located  close  together  for  ease  of  observation 
of  coincident  peaks  and  dips: 

1 —  PA  plate  current  and  power  output 

2 —  PA  r-f  grid  current  and  PA  plate 
current 

3 —  PA  r-f  grid  voltage  and  power  out- 
put 

4 —  Driver  plate  current  and  PA  r-f 
grid  voltage 

The  third  pair  listed  above  may  not  be 
necessary  if  the  PA  plate  current  dip  is  pro- 
nounced. When  this  instrumentation  is  pro- 
vided, the  neutralizing  procedure  is  as  follows: 


1— 

-Remove  the  r-f  feedback 

C6  j 

[f=' 

: 1 OPEN  1 

1 SHORT \ “ 

CHASSIS 

CC6 

CTX 

^ ' / //  /y.y///  / / / / z / / / . . .'1 

“Ct 

Figure  14 

FEEDBACK  SHORTING  DEVICE. 


2 —  Neutralize  the  grid-plate  capaci- 
tance of  the  driver  stage 

3 —  Neutralize  the  grid-plate  capaci- 
tance of  the  power  amplifier  (PA) 
stage 

4 —  Apply  r-f  feedback 

5 —  Neutralize  driver  grid-cathode  ca- 
pacitance 

These  steps  will  be  explained  in  more  detail 
in  the  following  paragraphs: 

Step  I.  The  removal  of  r-f  feedback  through 
the  feedback  circuit  must  be  complete.  The 
switch  (S>)  shown  in  the  feedback  circuit 
( figure  13)  is  one  satisfactory  method.  Since 
G,  is  effectively  across  the  PA  plate  tank  cir- 
cuit it  is  desirable  to  keep  it  across  the  circuit 
when  feedback  is  removed  to  avoid  appreciable 
detuning  of  the  plate  tank  circuit.  Another 
method  that  can  be  used  if  properly  done  is 
to  ground  the  junction  of  G and  G.  Ground- 
ing this  common  point  through  a switch  or 
relay  is  not  good  enough  because  of  common 
coupling  through  the  length  of  the  grounding 
lead.  The  grounding  method  shown  in  figure 
14  is  satisfactory. 

Step  2.  Plate  power  and  excitation  are  applied. 
The  driver  grid  tank  is  resonated  by  tuning 
for  a peak  in  driver  r-f  grid  voltage  or  driver 
plate  current.  The  power  amplifier  grid  tank 
circuit  is  then  resonated  and  adjusted  for  a 
dip  in  driver  plate  current.  Driver  neutraliza- 
tion is  now  adjusted  until  the  PA  r-f  grid 
voltage  (or  PA  grid  current)  peaks  at  exactly 
the  point  of  driver  plate  current  dip.  A handy 
rule  for  adjusting  grid-plate  neutralization  of 
a tube  without  feedback:  with  all  circuits  in 
resonance,  detune  the  plate  circuit  to  the  high 
frequency  side  of  resonance:  If  grid  current 
to  next  stage  (or  power  output  of  the  stage 
under  test)  increases,  more  neutralizing  capaci- 
tance is  required  and  vice  versa. 

If  the  driver  tube  operates  class  A so  that 
a plate  current  dip  cannot  be  observed,  a dif- 


HANDBOOK 


Neutralization  281 


R 


1 ^ 
K 4=09  Cio  ^ 

C__4__VLt . 

^ J 
— ^ C6 

11  • 

“C? 

Figure  15 

FEEDBACK  NEUTRALIZING 
CIRCUIT  USING 
AUXILIARY  RECEIVER. 


ferent  neutralizing  procedure  is  necessary  This 
will  be  discussed  in  a subsequent  section. 

Step  3.  This  is  the  same  as  step  2 except  it 
is  applied  to  the  power  amplifier  stage.  Ad- 
just the  neutralization  of  this  stage  for  a peak 
in  power  output  at  the  plate  current  dip. 

Step  4-  Reverse  step  1 and  apply  the  r-f  feed- 
back. 

Step  5>  Apply  plate  power  and  an  exciting  sig- 
nal to  drive  the  amplifier  to  nearly  full  out- 
put. Adjust  the  feedback  neutralization  for  a 
peak  in  amplifier  power  output  at  the  exact 
point  of  minimum  amplifier  plate  current. 
Decrease  the  feedback  neutralization  capaci- 
tance if  the  power  output  rises  when  the  tank 
circuit  is  tuned  to  the  high  frequency  side  of 
resonance. 

The  above  sequence  applies  when  the  neu- 
tralizing adjustments  are  approximately  cor- 
rect to  start  with.  If  they  are  far  off,  some  "cut- 
and-try”  adjustment  may  be  necessary.  Also, 
the  driver  stage  may  break  into  oscillation  if 
the  feedback  neutralizing  capacitance  is  not 
near  the  correct  setting. 

It  is  assumed  that  a single  tone  test  signal 
is  used  for  amplifier  excitation  during  the 
above  steps,  and  that  all  tank  circuits  are  at 
resonance  except  the  one  being  detuned  to 
make  the  observation.  There  is  some  interaction 
between  the  driver  neutralization  and  the  feed- 
back neutralization  so  if  an  appreciable  change 
is  made  in  any  adjustment  the  others  should 
be  rechecked.  It  is  important  that  the  grid-plate 
neutralization  be  accomplished  first  when  using 
the  above  procedure,  otherwise  the  feedback 
neutralization  will  be  off  a little,  since  it  par- 
tially compensates  for  that  error. 


Neulralizafion  The  method  of  neutralization 

Techniques  employing  a sensitive  r-f  de- 

tector inductively  coupled  to 
a tank  coil  is  difficult  to  apply  in  some  cases 
because  of  mechanical  construction  of  the 
equipment,  or  because  of  undesired  coupling. 
Another  method  for  observing  neutralization 
can  be  used,  which  appears  to  be  more  ac- 
curate in  actual  practice.  A sensitive  r-f  detec- 
tor such  as  a receiver  is  loosely  coupled  to  the 
grid  of  the  stage  being  neutralized,  as  shown 
in  figure  15-  The  coupling  capacitance  is  of 
the  order  of  one  or  two  ggfd.  It  must  be  small 
enough  to  avoid  upsetting  the  neutralization 
when  it  is  removed  because  the  total  grid- 
ground  capacitance  is  one  leg  of  the  neutraliz- 
ing bridge.  A signal  generator  is  connected  at 
point  S and  the  receiver  at  point  R.  If  Cm  is 
not  properly  adjusted  the  S-meter  on  the  re- 
ceiver will  either  kick  up  or  down  as  the  grid 
tank  circuit  is  tuned  through  resonance.  Cw 
may  be  adjusted  for  minimum  deflection  of  the 
S-meter  as  the  grid  circuit  is  tuned  through 
resonance. 

The  grid-plate  capacitance  of  the  tube  is 
then  neutralized  by  connecting  the  signal  gen- 
erator to  the  plate  of  the  tube  and  adjusting 
Cn  of  figure  1 3 for  minimum  deflection  again 
as  the  grid  tank  is  tuned  through  resonance. 
The  power  amplifier  stage  is  neutralized  in 
the  same  manner  by  connecting  a receiver 
loosely  to  the  grid  circuit,  and  attaching  a 
signal  generator  to  the  plate  of  the  tube.  The 
r-f  signal  can  be  fed  into  the  amplifier  output 
terminal  if  desired. 

Some  precautions  are  necessary  when  using 
this  neutralization  method.  First,  some  driver 
tubes  (the  6CL6,  for  example)  have  appre- 
ciably more  effective  input  capacitance  when 
in  operation  and  conducting  plate  current  than 
when  in  standby  condition.  This  increase  in 
input  capacitance  may  be  as  great  as  three  or 
four  ggfd,  and  since  this  is  part  of  the  neu- 
tralizing bridge  circuit  it  must  be  taken  into 
consideration.  The  result  of  this  change  in 
input  capacitance  is  that  the  neutralizing  ad- 
justment of  such  tubes  must  be  made  when 
they  are  conducting  normal  plate  current.  Stray 
coupling  must  be  avoided,  and  it  may  prove 
helpful  to  remove  filament  power  from  the 
preceding  stage  or  disable  its  input  circuit  in 
some  manner. 

It  should  be  noted  that  in  each  of  the  above 
adjustments  that  minimum  reaction  on  the 

grid  is  desired,  not  minimum  voltage.  Some 

residual  voltage  is  inherent  on  the  grid  when 
this  neutralizing  circuit  is  used. 


CHAPTER  FIFTEEN 


Amplitude  Modulation 


If  the  output  of  a c-w  transmitter  is  varied 
in  amplitude  at  an  audio  frequency  rate  in- 
stead of  interrupted  in  accordance  with  code 
characters,  a tone  will  be  heard  on  a receiver 
tuned  to  the  signal.  If  the  audio  signal  con- 
sists of  a band  of  audio  frequencies  com- 
prising voice  or  music  intelligence,  then  the 
voice  or  music  which  is  supetimposed  on  the 
radio  frequency  carrier  will  be  heard  on  the 
receiver. 

When  voice,  music,  video,  or  other  intelli- 
gence is  supetimposed  on  a radio  frequency 
carrier  by  means  of  a corresponding  variation 
in  the  amplitude  of  the  radio  frequency  output 
of  a transmitter,  amplitude  modulation  is  the 
result.  Telegraph  keying  of  a c-w  transmitter 
is  the  simplest  form  of  amplitude  modulation, 
while  video  modulation  in  a television  trans- 
mitter represents  a highly  complex  form.  Sys- 
tems for  modulating  the  amplitude  of  a carrier 
envelope  in  accordance  with  voice,  music,  or 
similar  rypes  of  complicated  audio  waveforms 
are  many  and  varied,  and  will  be  discussed 
later  on  in  this  chapter. 


15-1  Sidebands 

Modulation  is  essentially  a form  of  mixing 
or  combining  already  covered  in  a previous 
chapter.  To  transmit  voice  at  radio  frequencies 
by  means  of  amplitude  modulation,  the  voice 


frequencies  are  mixed  with  a radio  frequency 
carrier  so  that  the  voice  frequencies  are  con- 
verted to  radio  frequency  sidebands.  Though 
it  may  be  difficult  to  visualize,  the  amplitude 
of  the  radio  frequency  carrier  does  not  vary 
during  conventional  amplitude  modulation. 

Even  though  the  amplitude  of  radio  fre- 
quency voltage  representing  the  composite 
signal  ( resultant  of  the  carrier  and  sidebands, 
called  the  envelope)  will  vary  from  zero  to 
twice  the  unmodulated  signal  value  during 
full  modulation,  the  amplitude  of  the  carrier 
component  does  not  vary.  Also,  so  long  as 
the  amplitude  of  the  modulating  voltage  does 
not  vary,  the  amplitude  of  the  sidebands  will 
remain  constant.  For  this  to  be  apparent,  how- 
ever, it  is  necessary  to  measure  the  amplitude 
of  each  component  with  a highly  selective 
filter.  Otherwise,  the  measured  power  or  volt- 
age will  be  a resultant  of  two  or  more  of  the 
components,  and  the  amplitude  of  the  resultant 
will  vary  at  the  modulation  rate. 

If  a carrier  frequency  of  5000  kc.  is  modu- 
lated by  a pure  tone  of  1000  cycles,  or  1 kc., 
two  sidebands  are  formed:  one  at  5001  kc. 
(the  sum  frequency)  and  one  at  4999  kc.  (the 
difference  frequency).  The  frequency  of  each 
sideband  is  independent  of  the  amplitude  of 
the  modulating  tone,  or  modulation  percent- 
age; the  frequency  of  each  sideband  is  deter- 
mined only  by  the  frequency  of  the  modulat- 
ing tone.  This  assumes,  of  course,  that  the 
transmitter  is  not  modulated  in  excess  of  its 
linear  capability. 


282 


Modulation  283 


When  the  modulating  signal  consists  of 
multiple  frequencies,  as  is  the  case  with 
voice  or  music  modulation,  two  sidebands  will 
be  formed  by  each  modulating  frequency  (one 
on  each  side  of  the  carrier),  and  the  radiated 
signal  will  consist  of  a hand  of  frequencies. 
The  hand  width,  or  channel  taken  up  in  the 
frequency  spectrum  by  a conventional  double- 
sideband amplitude-modulated  signal,  is  equal 
to  twice  the  highest  modulating  frequency. 
For  example,  if  the  highest  modulating  fre- 
quency is  5000  cycles,  then  the  signal  (as- 
suming modulation  of  complex  and  varying 
waveform)  will  occupy  a band  extending  from 
5000  cycles  below  the  carrier  to  5000  cycles 
above  the  carrier. 

Frequencies  up  to  at  least  2500  cycles,  and 
preferably  3500  cycles,  are  necessary  for  good 
speech  intelligibility.  If  a filter  is  incorpo- 
rated in  the  audio  system  to  cut  out  all  fre- 
quencies above  approximately  3000.  cycles, 
the  band  width  of  a radio-telephone  signal  can 
be  limited  to  6 kc.  without  a significant  loss 
in  intelligibility.  However,  if  harmonic  distor- 
tion is  introduced  subsequent  to  the  filter,  as 
would  happen  in  the  case  of  an  overloaded 
modulator  or  overmodulation  of  the  carrier, 
new  frequencies  will  be  generated  and  the 
signal  will  occupy  a band  wider  than  6 kc. 


15-2  Mechanics  of 

Modulation 


A c-w  ot  unmodulated  t-f  carrier  wave  is 
represented  in  figure  lA.  An  audio  frequency 
sine  wave  is  represented  by  the  curve  of 
figure  IB.  When  the  two  are  combined  or 
"mixed,”  the  carrier  is  said  to  be  amplitude 
modulated,  and  a resultant  similar  to  1C  or 
ID  is  obtained.  It  should  be  noted  that  under 
modulation,  each  half  cycle  of  r-f  voltage 
differs  slightly  from  the  preceding  one  and 
the  following  one;  therefore  at  no  time  during 
modulation  is  the  r-f  waveform  a pure  sine 
wave.  This  is  simply  another  way  of  saying 
that  during  modulation,  the  transmitted  r-f 
energy  no  longer  is  confined  to  a single  radio 
frequency. 

It  will  be  noted  that  the  average  amplitude 
of  the  peak  r-f  voltage,  or  modulation  enve- 
lope, is  the  same  with  or  without  modulation. 
This  simply  means  that  the  modulation  is 
symmetrical  (assuming  a symmetrical  modu- 
lating wave)  and  that  for  distortionless  modu- 
lation the  upward  modulation  is  limited  to  a 
value  of  twice  the  unmodulated  carrier  wave 
amplitude  because  the  amplitude  cannot  go 
below  zero  on  downward  portions  of  the  mod- 
ulation cycle.  Figure  ID  illustrates  the  maxi- 


C.W. OR  UNMODULATED  CARRIER 


AUDIO  SIGNAL  FROM  MODULATOR 


50%  MODULATED  CARRIER 


100%  MODULATED  CARRIER 


Figure  1 

AMPLITUDE  MODULATED  WAVE 
Top  drawing  (A)  repreionfs  an  unmodulated 
carrier  wave;  (B)  shows  the  audio  output  of 
the  modulator.  Drawing  (C)  shows  the  audio 
signal  impressed  on  the  carrier  wave  to  the 
extent  of  50  per  cent  modulation;  (D)  shows 
the  carrier  with  100  per  cent  amplitude  modu- 
lation. 


mum  obtainable  distortionless  modulation  with 
a sine  modulating  wave,  the  r-f  voltage  at  the 
peak  of  the  r-f  cycle  varying  from  zero  to 
twice  the  unmodulated  vtilue,  and  the  r-f  power 
varying  from  zero  to  four  times  the  unmodu- 
lated value  ( the  power  varies  as  the  square 
of  the  voltage). 

While  the  average  r-f  voltage  of  the  modu- 
lated wave  over  a modulation  cycle  is  the 
same  as  for  the  unmodulated  carrier,  the  aver- 
age power  increases  with  modulation.  If  the 
radio  frequency  power  is  integrated  over  the 
audio  cycle,  it  will  be  found  with  100  pet  cent 
sine  wave  modulation  the  average  r-f  power 
has  increased  50  per  cent.  This  additional 
power  is  represented  by  the  sidebands,  be- 
cause as  previously  mentioned,  the  carrier 
power  does  not  vary  under  modulation.  Thus, 
when  a 100-watt  carrier  is  modulated  100  per 
cent  by  a sine  wave,  the  total  r-f  power  is  150 
watts;  100  watts  in  the  carrier  and  25  watts 
in  each  of  the  two  sidebands. 


284  Amplitude  Modulation 


THE  RADIO 


Modulation  So  long  as  the  relative  propor- 

Percentage  tion  of  the  various  sidebands 
making  up  voice  modulation  is 
maintained,  the  signal  may  be  received  and 
detected  without  distortion.  However,  the 
higher  the  average  amplitude  of  the  sidebands, 
the  greater  the  audio  signal  produced  at  the 
receiver.  For  this  reason  it  is  desirable  to 
increase  the  modulation  percentage,  or  degree 
of  modulation,  to  the  point  where  maximum 
peaks  just  hit  100  per  cent.  If  the  modulation 
percentage  is  increased  so  that  the  peaks  ex- 
ceed this  value,  distortion  is  introduced,  and 
if  carried  very  far,  bad  interference  to  signals 
on  nearby  channels  will  result. 

Modulation  The  amount  by  which  a carrier 
Measurement  is  being  modulated  may  be  ex- 
pressed either  as  a modulation 
factor,  varying  from  zero  to  1.0  at  maximum 
modulation,  or  as  a percentage.  The  percent- 
age of  modulation  is  equal  to  100  times  the 
modulation  factor.  Figure  2A  shows  a carrier 
wave  modulated  by  a sine-wave  audio  tone. 
A picture  such  as  this  might  be  seen  on  the 
screen  of  a cathode-ray  oscilloscope  with 
sawtooth  sweep  on  the  horizontal  plates  and 
the  modulated  carrier  impressed  on  the  verti- 
cal plates.  The  same  carrier  without  modulation 
would  appear  on  the  oscilloscope  screen  as 
figure  2B. 

The  percentage  of  modulation  of  the  posi- 
tive peaks  and  the  percentage  of  modulation 
of  the  negative  peaks  can  be  determined  sepa- 
rately from  two  oscilloscope  pictures  such 
as  shown. 

The  modulation  factor  of  the  positive  peaks 
may  be  determined  by  the  formula: 

Emax  ^cat 

M - 

F 

ar 

The  factor  for  negative  peaks  may  be  de- 
termined from  this  formula: 

F F • 

*-’cat  ^min 

M - 

E 

^car 

In  the  above  two  formulas  E^ax  max- 

imum carrier  amplitude  with  modulation  and 
Emin  is  the  minimum  amplitude;  Ecat  is  the 
steady-state  amplitude  of  the  carrier  with- 
out modulation.  Since  the  deflection  of  the 
spot  on  a cathode-ray  tube  is  linear  with  re- 
spect to  voltage,  the  relative  voltages  of 
these  various  amplitudes  may  be  determined 
by  measuring  the  deflections,  as  viewed  on 
the  screen,  with  a rule  calibrated  in  inches 
or  centimeters.  The  percentage  of  modulation 
of  the  carrier  may  be  had  by  multiplying  the 
modulation  factor  thus  obtained  by  100.  The 
above  procedure  assumes  that  there  is  no 


@ ® 


Figure  2 

GRAPHICAL  DETERMINATION  OF  MODU- 
LATION PERCENTAGE 

The  procedure  for  determining  modulation 
percentage  from  the  peak  voltage  points  in^ 
dicated  is  discussed  in  the  text. 


carrier  shift,  or  change  in  average  amplitude, 
with  modulation. 

If  the  modulating  voltage  is  symmetrical, 
such  as  a sine  wave,  and  modulation  is  ac- 
complished without  the  introduction  of  dis- 
tortion, then  the  percentage  modulation  will 
be  the  same  for  both  negative  and  positive 
peaks.  However,  the  distribution  and  phase 
relationships  of  harmonics  in  voice  and  music 
waveforms  are  such  that  the  percentage  modu- 
lation of  the  negative  modulation  peaks  may 
exceed  the  percentage  modulation  of  the  posi- 
tive peaks,  and  vice  versa.  The  percentage 
modulation  when  referred  to  without  regard 
to  polarity  is  an  indication  of  the  average  of 
the  negative  and  positive  peaks. 

Modulation  The  modulation  capability  of  a 
Capability  transmitter  is  the  maximum  per- 
centage to  which  that  transmitter 
may  be  modulated  before  spurious  sidebands 
ate  generated  in  the  output  or  before  the  dis- 
tortion of  the  modulating  waveform  becomes 
objectionable.  The  highest  modulation  cap- 
ability which  any  transmitter  may  have  on  the 
negative  peaks  is  100  per  cent.  The  maximum 
permissible  modulation  of  many  transmitters 
is  less  than  100  pet  cent,  especially  on  posi- 
tive peaks.  The  modulation  capability  of  a 
transmitter  may  be  limited  by  tubes  with  in- 
sufficient filament  emission,  by  insufficient 
excitation  or  grid  bias  to  a plate-modulated 
stage,  too  light  loading  of  any  type  of  ampli- 
fier carrying  modulated  r.f.,  insufficient  power 
output  capability  in  the  modulator,  or  too  much 
excitation  to  a grid-modulated  stage  or  a 
Class  B linear  amplifier.  In  any  case,  the 
FCC  regulations  specify  that  no  transmitter 
be  modulated  in  excess  of  its  modulation 
capability.  Hence,  it  is  desirable  to  make  the 
modulation  capability  of  a transmitter  as  near 
as  possible  to  100  pet  cent  so  that  the  cartiet 
power  may  be  used  most  effectively. 


HANDBOOK 


Modulation  Systems  285 


Speech  Waveform  The  manner  in  which  the 
Dissymmetry  human  voice  is  produced 

by  the  vocal  cords  gives 
rise  to  a certain  dissymmetry  in  the  waveform 
of  voice  sounds  when  they  are  picked  up  by 
a good-quality  microphone.  This  is  especially 
pronounced  in  the  male  voice,  and  more  so 
on  certain  voiced  sounds  than  on  others.  The 
result  of  this  dissymmetry  in  the  waveform  is 
that  the  voltage  peaks  on  one  side  of  the 
average  value  of  the  wave  will  be  consider- 
ably greater,  often  two  or  three  times  as  great, 
as  the  voltage  excursions  on  the  other  side 
of  the  zero  axis.  The  average  value  of  volt- 
age on  both  sides  of  the  wave  is,  of  course, 
the  same. 

As  a result  of  this  dissymmetry  in  the  male 
voice  waveform,  there  is  an  optimum  polarity 
of  the  modulating  voltage  that  must  be  ob- 
served if  maximum  sideband  energy  is  to  be 
obtained  without  negative  peak  clipping  and 
generation  of  "splatter”  on  adjacent  channels. 

A double-pole  double-throw  "phase  revers- 
ing” switch  in  the  input  or  output  leads  of  any 
transformer  in  the  speech  amplifier  system  will 
permit  poling  the  extended  peaks  in  the  direc- 
tion of  maximum  modulation  capability.  The 
optimum  polarity  may  be  determined  easily  by 
listening  on  a selective  receiver  tuned  to  a 
frequency  30  to  50  kc.  removed  from  the  de- 
sired signal  and  adjusting  the  phase  reversing 
switch  to  the  position  which  gives  the  least 
"splatter”  when  the  transmitter  is  modulated 
rather  heavily.  If  desired,  the  switch  then  may 
be  replaced  with  permanent  wiring,  so  long  as 
the  microphone  and  speech  system  are  not  to 
be  changed. 

A more  conclusive  illustration  of  the  lop- 
sidedness of  a speech  waveform  may  be  ob- 
tained by  observing  the  modulated  waveform 
of  a radiotelephone  transmitter  on  an  oscillo- 
scope. A portion  of  the  carrier  energy  of  the 
transmitter  should  be  coupled  by  means  of  a 
link  directly  to  the  vertical  plates  of  the 
’scope,  and  the  horizontal  sweep  should  be  a 
sawtooth  or  similar  wave  occurring  at  a rate 
of  approximately  30  to  70  sweeps  per  second. 

With  the  speech  signal  from  the  speech  am- 
plifier connected  to  the  transmitter  in  one  po- 
larity it  will  be  noticed  that  negative-peak 
clipping-as  indicated  by  bright  "spots”  in 
the  center  of  the  ’scope  pattern  whenever  the 
carrier  amplitude  goes  to  zero— will  occur  at 
a considerably  lower  level  of  average  modula- 
tion than  with  the  speech  signal  being  fed  to 
the  transmitter  in  the  other  polarity.  W'hen  the 
input  signal  to  the  transmitter  is  polarized  in 
such  a manner  that  the  "fingers”  of  the 
speech  wave  extend  in  the  direction  of  posi- 
tive modulation  these  fingers  usually  will  be 

clipped  in  the  plate  circuit  of  the  modulator 
at  an  acceptable  peak  modulation  level. 


The  use  of  the  proper  polarity  of  the  incom- 
ing speech  wave  in  modulating  a transmitter 
can  afford  an  increase  of  approximately  two 
to  one  in  the  amount  of  speech  audio  power 
which  may  be  placed  upon  the  carrier  for  an 
amplitude-modulated  transmitter  for  the  same 
amount  of  sideband  splatter.  More  effective 
methods  for  increasing  the  amount  of  audio 
power  on  the  carrier  of  an  AM  phone  trans- 
mitter are  discussed  later  in  this  chapter. 

Single-Sideband  Because  the  same  intelli- 

Tronsmission  gibility  is  contained  in  each' 

of  the  sidebands  associated 
with  a modulated  carrier,  it  is  not  necessary 
to  transmit  sidebands  on  both  sides  of  the 
carrier.  Also,  because  the  carrier  is  simply  a 
single  radio  frequency  wave  of  unvarying  am- 
plitude, it  is  not  necessary  to  transmit  the 
carrier  if  some  means  is  provided  for  inserting 
a locally  generated  carrier  at  the  receiver. 

Wlien  the  carrier  is  suppressed  but  both 
upper  and  lower  sidebands  are  transmitted,  it 
is  necessary  to  insert  a locally  generated 
carrier  at  the  receiver  of  exactly  the  same 
frequency  and  phase  as  the  carrier  which  was 
suppressed.  For  this  reason,  suppressed- 
carrier  double-sideband  systems  have  little 
practical  application. 

When  the  carrier  is  suppressed  and  only  the 
upper  or  the  lower  sideband  is  transmitted,  a 
highly  intelligible  signal  may  be  obtained  at 
the  receiver  even  though  the  locally  generated 
carrier  differs  a few  cycles  from  the  frequency 
of  the  carrier  which  was  suppressed  at  the 
transmitter.  A communications  system  utiliz- 
ing but  one  group  of  sidebands  with  carrier 
suppressed  is  known  as  a single  sideband 
system.  Such  systems  are  widely  used  for 
commercial  point  to  point  work,  and  are  being 
used  to  an  increasing  extent  in  amateur  com- 
munication. The  two  chief  advantages  of  the 
system  are:  ( 1)  an  effective  power  gain  of 
about  9 db  results  from  putting  all  the  radiat- 
ed power  in  intelligence  carrying  sideband 
frequencies  instead  of  mostly  into  radiated 
carrier,  and  (2)  elimination  of  the  selective 
fading  and  distortion  that  normally  occurs  in 
a conventional  double-sideband  system  when 
the  carrier  fades  and  the  sidebands  do  not,  or 
the  sidebands  fade  differently. 

15-3  Systems  of  Amplitude 
Modulation 


There  are  many  different  systems  and  meth- 
ods for  amplitude  modulating  a carrier,  but 
most  may  be  grouped  under  three  general  clas- 
sifications: (1)  variable  efficiency  systems 
in  which  the  average  input  to  the  stage  re- 


286  Amplitude  Modulation 


THE  RADIO 


mains  constant  with  and  without  modulation 
and  the  variations  in  the  efficiency  of  the 
stage  in  accordance  with  the  modulating  sig- 
nal accomplish  the  modulation;  (2)  constant 
efficiency  systems  in  which  the  input  to  the 
stage  is  varied  by  an  external  source  of  modu- 
lating energy  to  accomplish  the  modulation; 
and  (3)  so-called  high-efficiency  systems  in 
which  circuit  complexity  is  increased  to  ob- 
high  plate  circuit  efficiency  in  the  modulated 
stage  without  the  requirement  of  an  external 
high-level  modulator.  The  various  systems 
under  each  classification  have  individual 
characteristics  which  make  certain  ones  best 
suited  to  particular  applications. 

Variable  Efficiency  Since  the  average  input 

Modulation  remains  constant  in  a 

stage  employing  variable 
efficiency  modulation,  and  since  the  average 
power  output  of  the  stage  increases  with  modu- 
lation, the  additional  average  power  output 
from  the  stage  with  modulation  must  come  from 
the  plate  dissipation  of  the  tubes  in  the  stage. 
Hence,  for  the  best  relation  between  tube  cost 
and  power  output  the  tubes  employed  should 
have  as  high  a plate  dissipation  eating  per 
dollar  as  possible. 

The  plate  efficiency  in  such  an  amplifier  is 
doubled  when  going  from  the  unmodulated 
condition  to  the  peak  of  the  modulation  cycle. 
Hence,  the  unmodulated  efficiency  of  such  an 
amplifier  must  always  be  less  than  45  per 
cent,  since  the  maximum  peak  efficiency  ob- 
tainable in  a conventional  amplifier  is  in  the 
vicinity  of  90  per  cent.  Since  the  peak  effi- 
ciency in  certain  types  of  amplifiers  will  be 
as  low  as  60  per  cent,  the  unmodulated  effi- 
ciency in  such  amplifiers  will  be  in  the  vici- 
nity of  30  per  cent. 

Assuming  a typical  amplifier  having  a peak 
efficiency  of  70  per  cent,  the  following  fig- 
ures give  an  idea  of  the  operation  of  an  ideal- 
ized efficiency-modulated  stage  adjusted  for 
100  per  cent  sine-wave  modulation.  It  should 
be  kept  in  mind  that  the  plate  voltage  is  con- 
stant at  all  times,  even  over  the  audio  cycles. 


Plate  input  without  modulation 100  watts 

Output  without  modulation 35  watts 

Efficiency  without  modulation 35% 

Input  on  100%  positive  modulation 

peak  (plate  current  doubles) 200  watts 

Efficiency  on  100%  positive  peak....  70% 
Output  on  100%  positive  modula- 
tion peak 140  watts 

Input  on  100%  negative  peak 0 watts 

Efficiency  on  100%  negative  peak — 0% 

Output  on  100%  negative  pe^ 0 watts 


Average  input  with  100% 

modulation 100  watts 

Average  output  with  100%  modula- 
tion (35  watts  carrier  plus  17.5 

watts  sideband)  52.5  watts 

Average  efficiency  with  100% 

modulation 52.5% 

Systems  of  Efficiency  There  are  many  sys- 
Modulation  terns  of  efficiency  mod- 

ulation, but  they  all 
have  the  general  limitation  discussed  in  the 
previous  paragr^h— so  long  as  the  carrier 
amplitude  is  to  remain  constant  with  and 
without  modulation,  the  efficiency  at  carrier 
level  must  be  not  greater  than  one-half  the 
peak  modulation  efficiency  if  the  stage  is  to 
be  capable  of  100  pet  cent  modulation. 

The  classic  example  of  efficiency  modula- 
tion is  the  Class  B linear  r-f  amplifier,  to  be 
discussed  below.  The  other  three  common 
forms  of  efficiency  modulation  are  control- 
grid  modulation,  screen-grid  modulation,  and 
suppressor-grid  modulation.  In  each  case, 
including  that  of  the  Class  B linear  amplifier, 
note  that  the  modulation,  or  the  modulated 
signal,  is  impressed  on  a control  electrode 
of  the  stage. 

The  Class  B This  is  the  simplest  practi- 

Linear  Amplifier  cable  type  amplifier  for  an 
amplitude-modulated  wave 
or  a single-sideband  signal.  The  system  pos- 
sesses the  disadvantage  that  excitation,  grid 
bias,  and  loading  must  be  carefully  controlled 
to  preserve  the  linearity  of  the  stage.  Also, 
the  grid  circuit  of  the  tube,  in  the  usual  appli- 
cation where  grid  current  is  drawn  on  peaks, 
presents  a widely  varying  value  of  load  im- 
pedance to  the  source  of  excitation.  Hence  it 
is  necessary  to  include  some  sort  of  swamping 
resistor  to  reduce  the  effect  of  gtid-imped- 
ance  variations  with  modulation.  If  such  a 
swamping  resistance  across  the  grid  tank  is 
not  included,  or  is  too  high  in  value,  the  posi- 
tive modulation  peaks  of  the  incoming  modu- 
lated signal  will  tend  to  be  flattened  with 
resultant  distottion  of  the  wave  being  amplified. 

The  Class  B lineat  amplifier  has  long  been 
used  in  broadcast  transmitters,  but  recently 
has  received  much  more  general  usage  in  the 
h-f  range  for  two  significant  reasons:  (a)  the 
Class  B lineat  is  an  excellent  way  of  increas- 
ing the  powet  output  of  a single-sideband 
transmitter,  since  the  plate  efficiency  with 
full  signal  will  be  in  the  vicinity  of  70  per 
cent,  while  with  no  modulation  the  input  to 
the  stage  drops  to  a telativeiy  low  value;  and 
(b)  the  Class  B lineat  amplifier  operates  with 
relatively  low  harmonic  output  since  the  grid 
bias  on  the  stage  normally  is  slightly  less 


HANDBOOK 


Class  B Linear  Amplifier  287 


than  the  value  which  will  cut  off  plate  current 
to  the  stage  in  the  absence  of  excitation. 

Since  a Qass  B linear  amplifier  is  biased 
to  extended  cutoff  with  no  excitation  ( the 
grid  bias  at  extended  cutoff  will  be  approxi- 
mately equal  to  the  plate  voltage  divided  by 
the  amplification  factor  for  a triode,  and  will 
be  approximately  equal  to  the  scteen  voltage 
divided  by  the  grid-screen  mu  factor  for  a 
tetrode  or  pentode)  the  plate  current  will  flow 
essentially  in  180-degree  pulses.  Due  to  the 
relatively  large  operating  angle  of  plate  cur- 
rent flow  the  theoretical  peak  plate  efficiency 
is  limited  to  78.5  per  cent,  with  65  to  70  per 
cent  representing  a range  of  efficiency  nor- 
mally attainable,  and  the  harmonic  output 
will  be  low. 

The  carrier  power  output  from  a Class  B 
linear  amplifier  of  a normal  100  per  cent  mod- 
ulated AM  signal  will  be  about  one-half  the 
rated  plate  dissipation  of  the  stage,  with  opti- 
mum operating  conditions.  The  peak  output 
from  a Class  B linear,  which  represents  the 
maximum- signal  output  as  a single-sideband 
amplifier,  or  peak  output  with  a 100  per  cent 
AM  signal,  will  be  about  twice  the  plate  dis- 
sipation of  the  tubes  in  the  stage.  Thus  the 
carrier-level  input  power  to  a Class  B linear 
should  be  about  1.5  times  the  rated  plate  dis- 
sipation of  the  stage. 

The  schematic  circuit  of  a Class  B linear 
amplifier  is  the  same  as  a conventional  single- 
ended  or  push-pull  stage,  whether  triodes  or 
beam  tetrodes  are  used.  However,  a swamping 
resistor,  as  mentioned  before,  must  be  placed 
across  the  grid  tank  of  the  stage  if  the  oper- 
ating conditions  of  the  tube  are  such  that 
appreciable  grid  current  will  be  drawn  on  modu- 
lation peaks.  Also,  a fixed  source  of  grid  bias 
must  be  provided  for  the  stage.  A regulated 
grid-bias  power  supply  is  the  usual  source  of 
negative  bias  voltage. 

Adjustment  of  a Class  With  grid  bias  adjusted 
B L inear  Amplifier  to  the  correct  value, 

and  with  provision  for 
varying  the  excitation  voltage  to  the  stage 
and  the  loading  of  the  pl^te  circuit,  a fully 
modulated  signal  is  applied  to  the  grid  circuit 
of  the  stage.  Then  with  an  oscilloscope  cou- 
pled to  the  output  of  the  stage,  excitation  and 
loading  are  varied  until  the  stage  is  drawing 
the  normal  plate  input  and  the  output  wave- 
shape is  a good  replica  of  the  input  signal. 
The  adjustment  procedure  normally  will  re- 
quire a succession  of  approximations,  until 
the  optimum  set  of  adjustments  is  attained. 
Then  the  modulation  being  applied  to  the  in- 
put signal  should  be  removed  to  check  the 
linearity.  W'ith  modulation  removed,  in  the 
case  of  a 100  per  cent  AM  signal,  the  input 
to  the  stage  should  remain  constant,  and  the 


peak  output  of  the  r-f  envelope  should  fall  to 
half  the  value  obtained  on  positive  modula- 
tion peaks. 

Class  C . One  widely  used  system  of 

Grid  Modulation  efficiency  modulation  for 

communications  work  is 
Class  C control-grid  bias  modulation.  The  dis- 
tortion is  slightly  higher  than  for  a properly 
operated  Class  B linear  amplifier,  but  the  effi- 
ciency is  also  higher,  and  the  distortion  can 
be  kept  within  tolerable  limits  for  communi- 
cations work. 

Class  C grid  modulation  requites  high  plate 
voltage  on  the  modulated  stage,  if  maximum 
output  is  desired.  The  plate  voltage  is  nor- 
mally run  about  50  per  cent  higher  than  for 
maximum  output  with  plate  modulation. 

The  driving  power  required  for  operation  of 
a grid-modulated  amplifier  under  these  condi- 
tions is  somewhat  more  than  is  required  for 
operation  at  lower  bias  and  plate  voltage,  but 
the  increased  power  output  obtainable  over- 
balances the  additional  excitation  require- 
ment. Actually,  almost  half  as  much  excitation 
is  required  as  would  be  needed  if  the  same 
stage  were  to  be  operated  as  a Class  C plate- 
modulated  amplifier.  The  resistor  R across 
the  grid  tank  of  the  stage  serves  as  swamping 
to  stabilize  the  r-f  driving  voltage.  At  least 
50  per  cent  of  the  output  of  the  driving  stage 
should  be  dissipated  in  this  swamping  resistor 
under  carrier  conditions. 

A comparatively  small  amount  of  audio  power 
will  be  required  to  modulate  the  amplifier  stage 
100  pet  cent.  An  audio  amplifier  having  20 
watts  output  will  be  sufficient  to  modulate  an 
amplifier  with  one  kilowatt  input.  Proportion- 
ately smaller  amounts  of  audio  will  be  re- 
quired for  lower  powered  stages.  However,  the 
audio  amplifier  that  is  being  used  as  the  grid 
modulator  should,  in  any  case,  either  employ 
low  plate  resistance  tubes  such  as  2A3’s, 
employ  degenerative  feedback  from  the  output 
stage  to  one  of  the  preceding  stages  of  the 
speech  amplifier,  or  be  resistance  loaded  with 
a resistor  across  the  secondary  of  the  modu- 
lation transformer.  This  provision  of  low  driv- 
ing impedance  in  the  gridmodulator  is  to  insure 
good  regulation  in  the  audio  driver  for  the  grid 
modulated  stage.  Good  regulation  of  both  the 
audio  and  the  r-f  drivers  of  a grid-modulated 
stage  is  quite  important  if  distortion-free 
modulation  approaching  100  per  cent  is  desired, 
because  the  grid  impedance  of  the  modulated 
stage  varies  widely  over  the  audio  cycle. 

A practical  circuit  for  obtaining  grid-bias 
modulation  is  shown  in  figure  3.  The  modula- 
tor and  bias  regulator  tube  have  been  com- 
bined in  a single  6B4G  tube. 

The  regulator-modulator  tube  operates  as 
a cathode-follower.  The  average  d-c  voltage 


288  Amplitude  Modulation 


THE  RADIO 


R.F.  AMPLIFIER 


Figure  3 

GRID-BIAS  MODULATOR  CIRCUIT 


on  the  control  gtid  is  controlled  by  the  70,000- 
ohm  wire-wound  potentiometer  and  this  poten- 
tiometer adjusts  the  avetage  grid  bias  on  the 
modulated  stage.  However,  a-c  signal  voltage 
is  also  impressed  on  the  conttol-grid  of  the 
tube  and  since  the  cathode  follows  this  a-c 
wave  the  incoming  speech  wave  is  supetim- 
posed  on  the  avetage  gtid  bias,  thus  effecting 
grid-bias  modulation  of  the  r-f  amplifiet  stage. 
An  audio  voltage  swing  is  requited  on  the  grid 
of  the  6B4G  of  approximately  the  same  peak 
value  as  will  be  required  as  bias-voltage 
swing  on  the  grid-bias  modulated  stage.  This 
voltage  swing  will  normally  be  in  the  region 
from  50  to  200  peak  volts.  Up  to  about  100 
volts  peak  swing  can  be  obtained  from  a 6SJ7 
tube  as  a conventional  speech  amplifier  stage. 
The  higher  voltages  may  be  obtained  from  a 
tube  such  as  a 6J5  through  an  audio  trans- 
former of  2:1  or  2^^:1  ratio. 

With  the  normal  amount  of  comparatively 
tight  antenna  coupling  to  the  modulated  stage, 
a non-modulated  carrier  efficiency  of  40  per 
cent  can  be  obtained  with  substantially  dis- 
tortion-free modulation  up  to  practically  100 


per  cent.  If  the  antenna  coupling  is  decreased 

slightly  from  the  condition  just  described,  and 
the  excitation  is  increased  to  the  point  where 
the  amplifiet  draws  the  same  input,  carrier 
efficiency  of  50  per  cent  is  obtainable  with 
tolerable  distortion  at  90  per  cent  modulation. 

Tuning  the  The  most  satisfactory  pro- 

Grid-Bins  cedure  for  tuning  a stage 

Modulated  Stage  for  grid-bias  modulation  of 
the  Class  C type  is  as 
follows.  The  amplifier  should  first  be  neutra- 
lized, and  any  possible  tendency  toward  para- 
sitics  under  any  condition  of  operation  should 
be  eliminated.  Then  the  antenna  should  be 
coupled  to  the  plate  circuit,  the  grid  bias 
should  be  run  up  to  the  maximum  available 
value,  and  the  plate  voltage  and  excitation 
should  be  applied.  The  grid  bias  voltage 
should  then  be  reduced  until  the  amplifier 
draws  the  approximate  amount  of  plate  cur- 
rent it  is  desired  to  run,  and  modulation  corre- 
sponding to  about  80  pet  cent  is  then  applied, 
if  the  plate  current  kicks  up  when  modulation 
is  applied,  the  grid  bias  should  be  reduced; 
if  the  plate  meter  kicks  down,  increase  the 
grid  bias. 

When  the  amount  of  bias  voltage  has  been 
found  (by  adjusting  the  fine  control,  R2,  on 
the  bias  supply)  where  the  plate  meter  re- 
mains constant  with  modulation,  it  is  more 
than  probable  that  the  stage  will  be  drawing 
either  too  much  or  too  little  input.  The  an- 
tenna coupling  should  then  be  either  increased 
or  decreased  (depending  on  whether  the  in- 
put was  too  little  or  too  much,  respectively) 
until  the  input  is  more  nearly  the  correct  value. 
The  bias  should  then  be  readjusted  until  the 
plate  meter  remains  constant  with  modulation 
as  before.  By  slight  jockeying  back  and  forth 
of  antenna  coupling  and  grid  bias,  a point  can 
be  reached  where  the  tubes  are  running  at 
rated  plate  dissipation,  and  where  the  plate 
milliammeter  on  the  modulated  stage  remains 
substantially  constant  with  modulation. 

The  linearity  of  the  stage  should  then  be 
checked  by  any  of  the  conventional  methods; 
the  trapezoidal  pattern  method  employing  a 
cathode-ray  oscilloscope  is  probably  the  most 
satisfactory.  The  check  with  the  trapezoidal 
pattern  will  allow  the  determination  of  the 
proper  amount  of  gain  to  employ  on  the  speech 
amplifier.  Too  much  audio  power  on  the  grid 
of  the  modulated  stage  should  not  be  used  in 
the  tuning-up  process,  as  the  plate  meter  will 
kick  erratically  and  it  will  be  impossible  to 
make  a satisfactory  adjustment. 

Screen- Grid  Amplitude  modulation  may  be 

Modulation  accomplished  by  varying  the 

screen-grid  voltage  in  a Class 
C amplifier  which  employs  a pentode,  beam 


HANDBOOK 


Screen  Grid  Modulation 


289 


tetrode,  or  other  type  of  screen-grid  tube.  The 
modulation  obtained  in  this  way  is  not  es- 
pecially linear,  but  screen-grid  modulation 
does  offer  other  advantages  and  the  linearity 
is  quite  adequate  for  communications  work. 

There  are  two  significant  and  worthwhile 
advantages  of  screen-grid  modulation  for  com- 
munications work:  (1)  The  excitation  require- 
ments for  an  amplifier  which  is  to  be  modu- 
lated in  the  screen  are  not  at  all  critical,  and 
good  regulation  of  the  excitation  voltage  is 
not  required.  The  normal  rated  grid-circuit 
operating  conditions  specified  for  Class  C 
c-w  operation  are  quite  adequate  for  screen- 
grid  modulation.  (2)  The  audio  modulating 
power  requirements  for  screen-grid  modulation 
ate  relatively  low. 

A screen-grid  modulated  r-f  amplifier  oper- 
ates as  an  efficiency-modulated  amplifier,  the 
same  as  does  a Class  B linear  amplifier  and 
a grid-modulated  stage.  Hence,  plate  circuit 
loading  is  relatively  critical  as  in  any  effi- 
ciency-modulated stage,  and  must  be  adjusted 
to  the  correct  value  if  normal  power  output 
with  full  modulation  capability  is  to  be  ob- 
tained. As  in  the  case  of  any  efficiency-modu- 
lated stage,  the  operating  efficiency  at  the 
peak  of  the  modulation  cycle  will  be  between 
70  and  80  per  cent,  with  efficiency  at  the  car- 
rier level  ( if  the  stage  is  operating  in  the  nor- 
mal manner  with  full  carrier)  about  half  of  the 
peak-modulation  value. 

There  are  two  main  disadvantages  of  screen- 
grid  modulation,  and  several  factors  which 
must  be  considered  if  satisfactory  operation 
of  the  screen-grid  modulated  stage  is  to  be 
obtained.  The  disadvantages  are:  (1)  As  men- 
rioned  before,  the  linearity  of  modulation  with 
respect  to  screen-grid  voltage  of  such  a stage 
is  satisfactory  only  for  communications  work, 
unless  carrier-rectified  degenerative  feed-back 
is  employed  around  the  modulated  stage  to 
straighten  the  linearity  of  modulation.  ( 2)  The 
impedance  of  the  screen  grid  to  the  modulating 
signal  is  non-linear.  This  means  that  the  mod- 
ulating signal  must  be  obtained  from  a source 
of  quite  low  impedance  if  audio  distortion  of 
the  signal  appearing  at  the  screen  grid  is  to 
be  avoided. 

Screen-Grid  Instead  of  being  linear  with  re- 
Impedonce  spect  to  modulating  voltage,  as 
is  the  plate  circuit  of  a plate- 
modulated  Class  C amplifier,  the  screen  grid 
presents  approximately  a square-law  imped- 
ance to  the  modulating  signal  over  the  region 
of  signal  excursion  where  the  screen  is  posi- 
tive with  respect  to  ground.  This  non-linearity 
may  be  explained  in  the  following  manner:  At 

the  carrier  level  of  a conventional  screen- 

modulated  stage  the  plate-voltage  swing  of 
the  modulated  tube  is  one-half  the  voltage 


swing  at  peak-modulation  level.  This  condition 
must  exist  in  any  type  of  conventional  effi- 
ciency- modulated  stage  if  100  per  cent  posi- 
tive modulation  is  to  be  attainable.  Since  the 
plate-voltage  swing  is  at  half  amplitude,  and 
since  the  screen  voltage  is  at  half  its  full- 
modulation  value,  the  screen  current  is  rela- 
tively low.  But  at  the  positive  modulation  peak 
the  screen  voltage  is  approximately  doubled, 
and  the  plate-voltage  swing  also  is  at  twice 
the  carrier  amplitude.  Due  to  the  increase  in 
plate-voltage  swing  with  increasing  screen 
voltage,  the  screen  current  increases  more  than 
linearly  with  increasing  screen  voltage. 

In  a test  made  on  an  amplifier  with  an  813 
tube,  the  screen  current  at  carrier  level  was 
about  6 ma.  with  screen  potential  of  190  volts; 
but  under  conditions  which  represented  a posi- 
tive modulation  peak  the  screen  current  meas- 
ured 25  ma.  at  a potential  of  400  volts.  Thus 
instead  of  screen  current  doubling  with  twice 
screen  voltage  as  would  be  the  case  if  the 
screen  presented  a resistive  impedance,  the 
screen  current  became  about  four  times  as 
great  with  twice  the  screen  voltage. 

Another  factor  which  must  be  considered 
in  the  design  of  a screen-modulated  stage,  if 
full  modulation  is  to  be  obtained,  is  that  the 
power  output  of  a screen-grid  stage  with  zero 
screen  voltage  is  still  relatively  large.  Hence, 
if  anything  approaching  full  modulation  on 
negative  peaks  is  to  be  obtained,  the  screen 
potential  must  be  made  negative  with  respect 
to  ground  on  negative  modulation  peaks.  In 
the  usual  types  of  beam  tetrode  tubes  the 
screen  potential  must  be  20  to  50  volts  nega- 
tive with  respect  to  ground  before  cut-off  of 
output  is  obtained.  This  condition  further  com- 
plicates the  problem  ofobtaining  good  linearity 
in  the  audio  modulating  voltage  for  the  screen- 
modulated  stage,  since  the  screen  voltage 
must  be  driven  negatively  with  respect  to 
ground  over  a portion  of  the  cycle.  Hence  the 
screen  draws  no  current  over  a portion  of  the 
modulating  cycle,  and  over  the  major  portion 
of  the  cycle  when  the  screen  does  draw  cur- 
rent, it  presents  approximately  a square-law 
impedance. 

Circuits  for  Laboratory  analysis  of  a large 
Screon-Grid  number  of  circuits  for  accom- 

Modulation  plishing  screen  modulation  has 

led  to  the  conclusion  that  the 
audio  modulating  voltage  must  be  obtained 
from  a low-impedance  source  if  low-distor- 
tion modulation  is  to  be  obtained.  Figure  4 
shows  a group  of  sketches  of  the  modulation 
envelope  obtained  with  various  types  of  modu- 
lators and  also  with  insufficient  antenna  coup- 
ling. The  result  of  this  laboratory  work  led 
to  the  conclusion  that  the  cathode-follower 
modulator  of  the  basic  circuit  shown  in  figure 


290  Amplitude  Modulation 


the  radio 


ENVELOPE  OBTAINED  WITH 
INSUFFICIENT  ANTENNA 
COUPLING 


© 


© 


© 


Figure  4 

SCREEN-MODULATION  CIRCUITS 

Three  common  screen  moduiafion  circuits  ore  illustrated  above.  A//  three  circuits 
are  capable  of  giving  intelligible  voice  modulation  although  the  waveform  distortion 
in  the  circuits  of  (A)  end  (B)  Is  likely  to  be  rather  severe.  The  arrangement  at  (A) 
is  often  celled  ” clamp  tube"  screen  modulation;  by  returning  the  grid  leak  on  the 
clamp  tube  to  ground  the  circuit  will  give  controlled'-carrier  screen  modulation.  This 
circuit  has  the  advantage  that  it  Is  simple  and  is  well  suited  to  use  in  mobile  trans- 
mitters, (B)  is  an  arrangement  using  a trans^rmer  co^p/ec/  modulator,  and  offers  no 
particular  advantages.  The  arrangement  at  (C)  is  capable  of  giving  good  modulation 
linearity  due  to  the  low  impedance  of  the  cathode-follower  modulator.  However,  due 
to  the  relatively  low  heater-cathode  ratings  on  tubes  suited  for  use  os  the  modula- 
tor, a separate  heater  supply  for  the  modulator  tube  normally  is  required.  This  limi- 
tation makes  cpplication  of  the  circuit  to  the  mobile  transmitter  o special  problem, 
since  an  isolated  heater  supply  normally  is  not  available.  Shown  at  (D)  as  an  assist- 
ance in  the  tuning  of  a screen-modulated  transmitter  (or  any  efficiency-modulated 
transmitter  for  that  matter)  is  the  type  of  modulation  envelope  which  results  when 
loading  to  the  modulated  stage  is  insufficient. 


5 is  capable  of  giving  good-quality  screen- 
grid  modulation,  and  in  addition  the  circuit 
provides  convenient  adjustments  for  the  car- 
rier level  and  the  output  level  on  negative 
modulation  peaks.  This  latter  control,  P2  in 
figure  5,  allows  the  amplifier  to  be  adjusted 
in  such  a manner  that  negative-peak  clipping 
cannot  take  place,  yet  the  negative  modulation 
peaks  may  be  adjusted  to  a level  just  above 
that  at  which  sideband  splatter  will  occur. 

The  Cathode-  The  cathode  follower  is 

Follower  Modulator  ideally  suited  for  use  as 
the  modulator  for  a screen- 


grid  stage  since  it  acts  as  a relatively  low- 
impedance  source  of  modulating  voltage  for 
the  screen-grid  circuit.  In  addition  the  cathode- 
follower  modulator  allows  the  supply  voltage 
both  for  the  modulator  and  for  the  screen  grid 
of  the  modulated  tube  to  be  obtained  from  the 
high-voltage  supply  for  the  plate  of  the  screen- 
grid  tube  or  beam  tetrode.  In  the  usual  case 
the  plate  supply  for  the  cathode  follower,  and 
hence  for  the  screen  grid  of  the  modulated 
tube,  may  be  taken  from  the  bleeder  on  the 
high-voltage  power  supply.  A tap  on  the  bleeder 
may  be  used,  or  two  resistors  may  be  connect- 
ed in  series  to  make  up  the  bleeder,  with  ap- 


HANDBOOK 


Modulation  Systems 


291 


propriate  values  such  that  the  voltage  applied 
to  the  plate  of  the  cathode  follower  is  appro- 
priate for  the  tube  to  be  modulated.  It  is  im- 
portant that  a bypass  capacitor  be  used  from 
the  plate  of  the  cathode-follower  modulator 
to  ground. 

The  voltage  applied  to  the  plate  of  the 
cathode  follower  should  be  about  100  volts 
greater  than  the  rated  screen  voltage  for  the 
tetrode  tube  as  a c-w  Class  C amplifier.  Hence 
the  cathode-follower  plate  voltage  should  be 
about  350  volts  for  an  815,  2E26,  or  829B, 
about  400  volts  for  an  807  or  4-125A,  about 
500  volts  for  an  813,  and  about  600  volts  for 
a 4-250A  or  a 4E27.  Then  potentiometer  Pj 
in  figure  5 should  be  adjusted  until  the  carrier- 
level  screen  voltage  on  the  modulated  stage 
is  about  one-half  the  rated  screen  voltage 
specified  for  the  tube  as  a Class  C c-w  ampli- 
fier. The  current  taken  by  the  screen  of  the 
modulated  tube  under  carrier  conditions  will 
be  about  one-fourth  the  normal  screen  current 
for  c-w  operation. 

The  only  current  taken  by  the  cathode 
follower  itself  will  be  that  which  will  flow 
through  the  100,000-ohm  resistor  between  the 
cathode  of  the  6L6  modulator  and  the  nega- 
tive supply.  The  current  taken  from  the  bleeder 
on  the  high-voltage  supply  will  be  the  carrier- 
level  screen  current  of  the  tube  being  modu- 
lated (which  current  passes  of  course  through 
the  cathode  follower)  plus  that  current  which 
will  pass  through  the  100,000-ohm  resistor. 

The  loading  of  the  modulated  stage  should 
be  adjusted  until  the  input  to  the  tube  is  about 
50  per  cent  greater  than  the  rated  plate  dissi- 
pation of  the  tube  or  tubes  in  the  stage.  If  the 
carrier-level  screen  voltage  value  is  correct 
for  linear  modulation  of  the  stage,  the  loading 
will  have  to  be  somewhat  greater  than  that 
amount  of  loading  which  gives  maximum  output 
from  the  stage.  The  stage  may  then  be  modu- 
lated by  applying  an  audio  signal  to  the  grid 
of  the  cathode-follower  modulator,  while  ob- 
serving the  modulated  envelope  on  an  oscillo- 
scope. 

If  good  output  is  being  obtained,  and  the 
modulation  envelope  appears  as  shown  in  fig- 
ure 4C,  all  is  well,  except  that  P2  in  figure  5 
should  be  adjusted  until  negative  modulation 
peaks,  even  with  excessive  modulating  signal, 
do  not  cause  carrier  cutoff  with  its  attendant 
sideband  splatter.  If  the  envelope  appears  as 
at  figure  4D,  antenna  coupling  should  be  in- 
creased while  the  carrier  level  is  backed  down 
by  potentiometer  Pj  in  figure  5 until  a set  of 
adjustments  is  obtained  which  will  give  a satis- 
factory modulation  envelope  as  shown  in 
figure  4C. 

Changing  Bands  After  a satisfactory  set  of  ad- 
justments has  been  obtained. 


Figure  5 

CATHODE-FOLLOWER 
SCREEN-MODULATION  CIRCUIT 
A deta/fed  discussion  of  this  circuit,  which 
also  is  represented  in  figure  4C,  is  given  in 
the  accompanying  text. 


it  is  not  difficult  to  readjust  the  amplifier  for 
operation  on  different  bands.  Potentiometers 
Pj  (carrier  level),  and  P2  (negative  peak  level) 
may  be  left  fixed  after  a satisfactory  adjust- 
ment, with  the  aid  of  the  scope,  has  once  been 
found.  Then  when  changing  bands  it  is  only 
necessary  to  adjust  excitation  until  the  correct 
value  of  grid  current  is  obtained,  and  then  to 
adjust  antenna  coupling  until  correct  plate 
current  is  obtained.  Note  that  the  correct  plate 
current  for  an  efficiency-modulated  amplifier 
is  only  slightly  less  than  the  out-of-resonance 
plate  current  of  the  stage.  Hence  carrier-level 
screen  voltage  must  be  low  so  that  the  out-of- 
resonance plate  current  will  not  be  too  high, 
and  relatively  heavy  antenna  coupling  must  be 
used  so  that  the  operating  plate  current  will 
be  near  the  out-of-resonance  value,  and  so  that 
the  operating  input  will  be  slightly  greater 
than  1.5  times  the  rated  plate  dissipation  of 
the  tube  or  tubes  in  the  stage.  Since  the  carrier 
efficiency  of  the  stage  will  be  only  35  to  40 
per  cent,  the  tubes  will  be  operating  with  plate 
dissipation  of  approximately  the  rated  value 
without  modulation. 

Speech  Clipping  in  The  maximum  r-f  output 
the  Modulated  Stage  of  an  efficiency-modu- 
lated stage  is  limited 
by  the  maximum  possible  plate  voltage  swing 
on  positive  modulation  peaks.  In  the  modula- 
lation  circuit  of  figure  5 the  minimum  output 
is  limited  by  the  minimum  voltage  which  the 
screen  will  reach  on  a negative  modulation 
peak,  as  set  by  potentiometer  P2.  Hence  the 
screen-grid-modulated  stage,  when  using  the 
modulator  of  figure  5,  acts  effectively  as  a 
speech  clipper,  provided  the  modulating  signal 
amplitude  is  not  too  much  more  than  that  v^ue 


292  Amplitude  Modulation 


THE  RADIO 


which  will  accomplish  full  modulation.  With 

correct  adjustments  of  the  operating  conditions 
of  the  stage  it  can  be  made  to  clip  positive 
and  negative  modulation  peaks  symmetrically. 
However,  the  inherent  peak  clipping  ability  of 
the  stage  should  not  be  relied  upon  as  a means 
of  obtaining  a large  amount  of  speech  com- 
pression, since  excessive  audio  distortion  and 
excessive  screen  current  on  the  modulated 
stage  will  result. 

Characteristics  of  a An  important  character- 
Typical  Screen-  istic  of  the  screen-modu- 

Modulated  Stage  lated  stage,  when  using 

the  cathode-follower  mod- 
ulator, is  that  excessive  plate  voltage  on  the 
modulated  stage  is  not  required.  In  fact,  full 
output  usually  may  be  obtained  with  the  larger 
tubes  at  an  operating  plate  voltage  from  one- 
half  to  two-thirds  the  maximum  rated  plate 
voltage  for  c-w  operation.  This  desirable  con- 
dition is  the  natural  result  of  using  a low- 
impedance  source  of  modulating  signal  for 
the  stage. 

As  an  example  of  a typical  screen-modu- 
lated stage,  full  output  of  75  watts  of  carrier 
may  be  obtained  from  an  813  tube  operating 
with  a plate  potential  of  only  1250  volts.  No 
increase  in  output  from  the  813  may  be  ob- 
tained by  increasing  the  plate  voltage,  since 
the  tube  may  be  operated  with  full  rated  plate 
dissipation  of  125  watts,  with  normal  plate 
efficiency  for  a screen-modulated  stage,  37.5 
per  cent,  at  the  1250-volt  potential. 

The  operating  conditions  of  a screen-modu- 
lated 813  stage  are  as  follows: 

Plate  voltage— 1250  volts 
Plate  current  — 160  ma. 

Plate  input— 200  watts 
Grid  current— 11  ma. 

Grid  bias— 110  volts 

Carrier  screen  voltage— 190  volts 

Carrier  screen  current— 6 ma. 

Power  output— approx.  75  watts 

With  full  100  per  cent  modulation  the  plate 
current  decreases  about  2 ma.  and  the  screen 
current  increases  about  1 ma. ; hence  plate, 
screen,  and  grid  current  remain  essentially 
constant  with  modulation.  Referring  to  figure 
5,  which  was  the  circuit  used  as  modulator 
for  the  813,  (El)  measured  plus  155  volts,  ( E2) 
measured  —50  volts,  ( E3)  measured  plus  190 
volts,  ( E4)  measured  plus  500  volts,  and  the 
r.m. s.  swing  at  (E5)  for  full  modulation  meas- 
ured 210  volts,  which  represents  a peak  swing 
of  about  296  volts.  Due  to  the  high  positive 
voltage,  and  the  large  audio  swing,  on  the 
cathode  of  the  6L6  (triode  connected)  modu- 
lator tube,  it  is  important  that  the  heater  of 
of  this  tube  be  fed  from  a separate  filament 


transformer  or  filament  winding.  Note  also  that 

the  operating  plate-to-cathode  voltage  on  the 
6L6  modulator  tube  does  not  exceed  the  360- 
volt  rating  of  the  tube,  since  the  operating 
potential  of  the  cathode  is  considerably  above 
ground  potential. 

Suppressor- Grid  Still  another  form  of  effi- 
Modulation  ciency  modulation  may  be 

obtained  by  applying  the 
audio  modulating  signal  to  the  suppressor  grid 
of  a pentode  Class  C r-f  amplifier.  Basically, 
suppressor- grid  modulation  operates  in  the 
same  general  manner  as  other  forms  of  effi- 
ciency modulation;  carrier  plate  circuit  effi- 
ciency is  about  35  per  cent,  and  antenna  coup- 
ling must  be  rather  tight.  However,  suppressor- 
grid  modulation  has  one  sizeable  disadvantage, 
in  addition  to  the  fact  that  pentode  tubes  are 
not  nearly  so  widely  used  as  beam  tetrodes 
which  of  course  do  not  have  the  suppressor 
element.  This  disadvantage  is  that  the  screen- 
grid  current  to  a suppressor-grid  modulated 
amplifier  is  rather  high.  The  high  screen  cur- 
rent is  a natural  consequence  of  the  rather  high 
negative  bias  on  the  suppressor  grid,  which 
reduces  the  plate-voltage  swing  and  plate  cur- 
rent with  a resulting  increase  in  the  screen 
current. 

In  tuning  a suppressor-grid  modulated  am- 
plifier, the  grid  bias,  grid  current,  screen  volt- 
age, and  plate  voltage  are  about  the  same  as 
for  Class  C c-w  operation  of  the  stage.  But 
the  suppressor  grid  is  biased  negatively  to  a 
value  which  reduces  the  plate-circuit  effici- 
ency to  about  one-half  the  maximum  obtainable 
from  the  particular  amplifier,  with  antenna 
coupling  adjusted  until  the  plate  input  is  about 
1.5  times  the  rated  plate  dissipation  of  the 
stage.  It  is  important  that  the  input  to  the 
screen  grid  be  measured  to  make  sure  that  the 
rated  screen  dissipation  of  the  tube  is  not 
being  exceeded.  Then  the  audio  signal  is  ap- 
plied to  the  suppressor  grid.  In  the  normal 
application  the  audio  voltage  swing  on  the 
suppressor  will  be  somewhat  greater  than  the 
negative  bias  on  the  element.  Hence  sup- 
pressor-grid current  will  flow  on  modulation 
peaks,  so  that  the  source  of  audio  signal  volt- 
age must  have  good  regulation.  Tubes  suitable 
for  suppressor-grid  modulation  are:  2E22, 
837,  4E27/8001,  5-125,  804  and  803.  A typi- 
cal suppressor-grid  modulated  amplifier  is 
illustrated  in  figure  6. 


15-4  Input  Modulation 

Systems 

Constant  efficiency  variable-input  modula- 
tion systems  operate  by  virtue  of  the  addition 


HANDBOOK 


Plate  Modulation  293 


Figure  6 

AMPLIFIER  WITH  SUPPRESSOR-GRID 
MODULATION 

Recomm&nded  operating  conditions  for  tin- 
ear  suppressor-grid  modulation  of  a 4E27/ 
2578/800  7 stage  are  given  on  the  drawing- 


of  external  power  to  the  modulated  stage  to 
effect  the  modulation.  There  are  two  general 
classifications  that  come  under  this  heading; 
those  systems  in  which  the  additional  power 
is  supplied  as  audio  frequency  energy  from  a 
modulator,  usually  called  plate  modulation 
systems,  and  those  systems  in  which  the  addi- 
tional power  to  effect  modulation  is  supplied 
as  direct  current  from  the  plate  supply. 

Under  the  former  classiUcation  comes  Heis- 
ing  modulation  (probably  the  oldest  type  of 
modulation  to  be  applied  to  a continuous  car- 
rier), Class  B plate  modulation,  and  series 
modulation.  These  types  of  plate  modulation 
are  by  far  the  easiest  to  get  into  operation, 
and  they  give  a very  good  ratio  of  power  input 
to  the  modulated  stage  to  power  output;  65  to 
80  per  cent  efficiency  is  the  general  rule.  It 
is  for  these  two  important  reasons  that  these 
modulation  systems,  particularly  Class  B plate 
modulation,  are  at  present  the  most  popular 
for  communications  work. 

Modulation  systems  coming  under  the  sec- 
ond classification  are  of  comparatively  recent 
development  but  have  been  widely  applied  to 
broadcast  work.  There  are  quite  a few  systems 
in  this  class.  Two  of  the  more  widely  used 
are  the  Doherty  linear  amplifier,  and  the  Ter- 
man-Woodyard  high-efficiency  grid-modulated 
amplifier.  Both  systems  operate  by  virtue  of 
a carrier  amplifier  and  a peak  amplifier  con- 
nected together  by  electrical  quarter-wave 
lines.  They  will  be  described  later  in  this 
section. 

Plate  Madulatian  Plate  modulation  is  the  ap- 
plication of  the  audio  power 


to  the  plate  circuit  of  an  r-f  amplifier.  The  r-f 
amplifier  must  be  operated  Class  C for  this 
type  of  modulation  in  order  to  obtain  a radio- 
frequency  output  which  changes  in  exact  ac- 
cordance with  the  variation  in  plate  voltage. 
The  r-f  amplifier  is  100  per  cent  modulated 
when  the  peak  a-c  voltage  from  the  modulator 
is  equal  to  the  d.c.  voltage  applied  to  the  r-f 
tube.  The  positive  peaks  of  audio  voltage  in- 
crease the  instantaneous  plate  voltage  on  the 
r-f  tube  to  twice  the  d-c  value,  and  the  nega- 
tive peaks  reduce  the  voltage  to  zero. 

The  instantaneous  plate  current  to  the  r-f 
stage  also  varies  in  accordance  with  the  modu- 
lating voltage.  The  peak  alternating  current 
in  the  output  of  a modulator  must  be  equal  to 
the  d-c  plate  current  of  the  Class  C r-f  stage 
at  the  point  of  100  per  cent  modulation.  This 
combination  of  change  in  audio  voltage  and 
current  can  be  most  easily  referred  to  in  terms 
of  audio  power  in  watts. 

In  a sinusoidally  modulated  wave,  the  an- 
tenna current  increases  approximately  22  pet 
cent  for  100  per  cent  modulation  with  a pure 
tone  input;  an  r-f  meter  in  the  antenna  circuit 
indicates  this  increase  in  antenna  current. 
The  average  power  of  the  r-f  wave  increases 
50  pet  cent  for  100  per  cent  modulation,  the 
efficiency  remaining  constant. 

This  indicates  that  in  a plate-modulated 
radiotelephone  transmitter,  the  audio-frequency 
channel  must  supply  this  additional  50  per 
cent  increase  in  average  power  for  sine-wave 
modulation.  If  the  power  input  to  the  modu- 
lated stage  is  100  watts,  for  example,  the 
average  power  will  increase  to  150  watts  at 
100  per  cent  modulation,  and  this  additional 
50  watts  of  power  must  be  supplied  by  the 
modulator  when  plate  modulation  is  used.  The 
actual  antenna  power  is  a constant  percentage 
of  the  total  value  of  input  power. 

One  of  the  advantages  of  plate  (or  power) 
modulation  is  the  ease  with  which  proper  ad- 
justments can  be  made  in  the  transmitter.  Also, 
there  is  less  plate  loss  in  the  r-f  amplifier  for 
a given  value  of  carrier  power  than  with  other 
forms  of  modulation  because  the  plate  effi- 
ciency is  higher. 

By  properly  matching  the  plate  impedance 
of  the  r-f  tube  to  the  output  of  the  modulator, 
the  ratio  of  voltage  and  current  swing  to  d-c 
voltage  and  current  is  automatically  obtained. 
The  modulator  should  have  a peak  voltage 
output  equal  to  the  average  d-c  plate  voltage 
on  the  modulated  stage.  The  modulator  should 
also  have  a peak  power  output  equal  to  the 
d-c  plate  input  power  to  the  modulated  stage. 

The  average  power  output  of  the  modulator  will 
depend  upon  the  type  of  waveform.  If  the  am- 
plifier is  being  Heising  modulated  by  a Class 
A stage,  the  modulator  must  have  an  average 


294  Amplitude  Modulation 


THE  RADIO 


MODULATED  CLASS  C 

R.F.  AMPLIFIER 


Figure  7 

HEISING  PLATE  MODULATION 


This  type  of  modulation  was  the  first  form 
of  plate  modulation.  It  is  sometimes  known 
as  constant  current**  modulation.  Because 
of  the  effective  1:1  ratio  of  the  coupling 
choke,  it  is  impossible  to  obtain  100  per  cent 
modulation  unless  the  plate  voltage  to  the 
modulated  stage  is  dropped  slightly  by  re- 
sistor  R.  The  capacitor  C merely  bypasses 
the  audio  around  R,  so  that  the  full  o-f  out* 
put  voltage  of  the  modulator  is  Impressed 
on  the  Class  C stage. 


CLASS  C 
AMPLIFIER 


Figure  8 

CLASS  B PLATE  MODULATION 


Thii  type  of  modulaflon  li  the  most  flexible 
in  that  the  loading  adjustment  can  be  made 
In  a short  period  of  time  and  without  elabo- 
rate test  equipment  after  a change  in  oper- 
ating frequency  of  the  Class  C amplifier  has 
been  made. 


power  output  capability  of  one-half  the  input 
to  the  Class  C stage.  If  the  modulator  is  a 
Class  B audio  amplifier,  the  average  power 
required  of  it  may  vary  from  one-quarter  to  more 
than  one-half  the  Class  C input  depending 
upon  the  waveform.  However,  the  peak  power 
output  of  any  modulator  must  be  equal  to  the 
Class  C input  to  be  modulated. 

Heising  Heising  modulation  is  the  oldest 

Modulation  system  of  plate  modulation,  and 
usually  consists  of  a Class  A 
audio  amplifier  coupled  to  the  r-f  amplifier  by 
means  of  a modulation  choke  coil,  as  shown 
in  figure  7. 

The  d.c.  plate  voltage  and  plate  current  in 
the  r-f  amplifier  must  be  adjusted  to  a value 
which  will  cause  the  plate  impedance  to  match 
the  output  of  the  modulator,  since  the  modula- 
tion choke  gives  a 1-to-l  coupling  ratio.  A 
series  resistor,  by-passed  for  audio  frequen- 
cies by  means  of  a capacitor,  must  be  connect- 
ed in  series  with  the  plate  of  the  r-f  amplifier 
to  obtain  modulation  up  to  100  per  cent.  The 
peak  output  voltage  of  a Class  A amplifier 
does  not  reach  a value  equal  to  the  d-c  voltage 
tq>plied  to  the  amplifier  and,  consequently, 
the  d-c  plate  voltage  impressed  across  the 
r-f  tube  must  be  reduced  to  a value  equal  to 


the  maximum  available  a-c  peak  voltage  if 
100%  modulation  is  to  be  obtained. 

A higher  degree  of  distortion  can  be  toler- 
ated in  low-power  emergency  phone  transmitters 
which  use  a pentode  modulator  tube,  and  the 
series  resistor  and  by-pass  capacitor  are 
usually  omitted  in  such  transmitters. 

Class  B High-level  Class  B plate 

Plate  Madulatlon  modulation  is  the  least  ex- 
pensive method  of  plate 
modulation.  Figure  8 shows  a conventional 
Class  B plate-modulated  Class  C amplifier. 

The  statement  that  the  modulator  output 
power  must  be  one-half  the  Class  C input  for 
100  per  cent  modulation  is  correct  only  if  the 
waveform  of  the  modulating  power  is  a sine 
wave.  Where  the  modulator  waveform  is  un- 
dipped speech,  the  average  modulator  power 
for  100  per  cent  modulation  is  considerably 
less  than  one-half  the  Class  C input. 

Powar  Relations  in  It  has  been  determined  ex- 
Speech  Waveforms  perimentally  that  the  ratio 
of  peak  to  average  power 
in  a speech  waveform  is  approximately  4 to  1 
as  contrasted  to  a ratio  of  2 to  1 in  a sine 
wave.  This  is  due  to  the  high  harmonic  con- 
tent of  such  a waveform,  and  to  the  fact  that 


HANDBOOK 


Plate  Modulation  2 95 


this  high  harmonic  content  manifests  itself  by 
making  the  wave  unsymmetrical  and  causing 
sharp  peaks  or  "fingers”  of  high  energy  con- 
tent to  appear.  Thus  for  undipped  speech,  the 
average  modulator  plate  current,  plate  dissi- 
pation, and  power  output  are  approximately 
one-half  the  sine  wave  values  for  a given  peak 
output  power. 

Both  peak  power  and  average  power  are 
necessarily  associated  with  waveform.  Peak 
power  is  just  what  the  name  implies;  the  power 
at  the  peak  of  a wave.  Peak  power,  although 
of  the  utmost  importance  in  modulation,  is  of 
no  great  significance  in  a-c  power  work,  ex- 
cept insofar  as  the  average  power  may  be  de- 
termined from  the  peak  value  of  a known  wave 
form. 

There  is  no  time  element  implied  in  the 
definition  of  peak  power;  peak  power  may  be 
instantaneous— and  for  this  reason  average 
power,  which  is  definitely  associated  with 
time,  is  the  important  factor  in  plate  dissipa- 
tion. It  is  possible  that  the  pe^  power  of  a 
given  waveform  be  several  times  the  average 
value;  for  a sine  wave,  the  peak  power  is  twice 
the  average  value,  and  for  undipped  speech 
the  peak  power  is  approximately  four  times 
the  average  value.  For  100  per  cent  modula- 
tion, the  peak  (instantaneous)  audio  power 
must  equal  the  Class  C input,  although  the 
average  power  for  this  value  of  peak  varies 
widely  depending  upon  the  modulator  wave- 
form, being  greater  than  50  per  cent  for  speech 
that  has  been  clipped  and  filtered,  50  per  cent 
for  a sine  wave,  and  about  25  pet  cent  for  typ- 
ical undipped  speech  tones. 

Modulation  The  modulation  transformer  is 
Transformer  a device  for  matching  the  load 
Calculations  impedance  of  the  Class  C am- 
plifier to  the  recommended  load 
impedance  of  the  Class  B modulator  tubes. 
Modulation  transformers  intended  for  com- 
munications work  are  usually  designed  to 
carry  the  Class  C plate  current  through  their 
secondary  windings,  as  shown  in  figure  8. 
The  manufacturer’s  ratings  should  be  con- 
sulted to  insure  that  the  d-c  plate  current 
passed  through  the  secondary  winding  does 
not  exceed  the  maximum  rating. 

A detailed  discussion  of  the  method  of 
making  modulation  transformer  calculations 
has  been  given  in  Chapter  Six.  However,  to 
emphasize  the  method  of  making  the  calcula- 
tion, an  additional  example  will  be  given. 

Suppose  we  take  the  case  of  a Class  C am- 
plifier operating  at  a plate  voltage  of  2000 
with  225  ma.  of  plate  current.  This  amplifier 
would  present  a load  resistance  of  2000  divi- 
ded by  0.225  amperes  or  8888  ohms.  The  plate 
power  input  would  be  2000  times  0.225  or  450 
watts.  By  reference  to  Chapter  Six  we  see  that 


a pair  of  811  tubes  operating  at  1500  plate 
volts  will  deliver  225  watts  of  audio  output. 
The  plate-to-plate  load  resistance  for  these 
tubes  under  the  specified  operating  conditions 
is  18,000  ohms.  Hence  our  problem  is  to  match 
the  Class  C amplifier  load  resistance  of  8888 
ohms  to  the  18,000-ohm  load  resistance  re- 
quired by  the  modulator  tubes. 

A 200-to-300  watt  modulation  transformer 
will  be  required  for  the  job.  If  the  taps  on  the 
transformer  are  given  in  terms  of  impedances 
it  will  only  be  necessary  to  connect  the  sec- 
ondary for  8888  ohms  (or  a value  approximately 
equal  to  this  such  as  9000  ohms)  and  the  pri- 
mary for  18,000  ohms.  If  it  is  necessary  to 
determine  the  proper  turns  ratio  required  of  the 
transformer  it  can  be  determined  in  the  follow- 
ing manner.  The  square  root  of  the  impedance 
ratio  is  equal  to  the  turns  ratio,  hence; 

8888  = V 0.494  = 0.703 
V 18000 

The  transformer  must  have  a turns  ratio  of 
approximately  l-to-0.7  step  down,  total  pri- 
mary to  total  secondary.  The  greater  number 
of  turns  always  goes  with  the  higher  imped- 
ance, and  vice  versa. 

Plote-and-Screen  When  only  the  plate  of  a 
Modulation  screen-grid  tube  is  modu- 

lated, it  is  impossible  to  ob- 
tain high-percentage  linear  modulation  under 
ordinary  conditions.  The  plate  current  of  such 
a stage  is  not  linear  with  plate  voltage.  How- 
ever, if  the  screen  is  modulated  simultaneously 
with  the  plate,  the  instantaneous  screen  volt- 
age drops  in  proportion  to  the  drop  in  the  plate 
voltage,  and  linear  modulation  can  then  be  ob- 
tained. Four  satisfactory  circuits  for  accom- 
plishing combined  plate  and  screen  modula- 
tion are  shown  in  figure  9- 

The  screen  r-f  by-pass  capacitor  C2,  should 
not  have  a greater  value  than  0.005  pid.,  pref- 
erably not  larger  than  0.001  pfd.  It  should  be 
large  enough  to  bypass  effectively  all  r-f  volt- 
age without  short-circuiting  high-frequency 
audio  voltages.  The  plate  by-pass  capacitor 
can  be  of  any  value  from  0.  002  pfd.  to  0.  005 
pfd.  The  screen-dropping  resistor,  should 
reduce  the  applied  high  voltage  to  the  value 
specified  for  operating  the  particular  tube  in 
the  circuit.  Capacitor  Cj  is  seldom  required 
yet  some  tubes  may  require  this  capacitor  in 
order  to  keep  C2  from  attenuating  the  high  fre- 
quencies. Different  values  between  .0002  and 
.002  ftfd.  should  be  tried  for  best  results. 

Figure  9C  shows  another  method  which  uses 
a third  winding  on  the  modulation  transformer, 
through  which  the  screen-grid  is  connected  to 


2 96  Amplitude  Modulation 


THE  RADIO 


a low-voltage  power  supply.  The  ratio  of  turns 
between  the  two  output  windings  depends  upon 
the  type  of  screen-grid  tube  which  is  being 
modulated.  Normally  it  will  be  such  that  the 
screen  voltage  is  being  modulated  60  pet  cent 
when  the  plate  voltage  is  receiving  100  pet 
cent  modulation. 

If  the  screen  voltage  is  derived  from  a drop- 
ping resistor  ( not  a divider)  that  is  bypassed 
for  r.f.  but  not  a.f.,  it  is  possible  to  secure 
quite  good  modulation  by  applying  modulation 
only  to  the  plate.  Under  these  conditions,  the 
screen  tends  to  modulate  itself,  the  screen 
voltage  varying  over  the  audio  cycle  as  a re- 
sult of  the  screen  impedance  increasing  with 
plate  voltage,  and  decreasing  with  a decrease 


in  plate  voltage.  This  circuit  arrangement  is 
illustrated  in  figure  9B. 

A similar  application  of  this  principle  is 
shown  in  figure  9D.  In  this  case  the  screen 
voltage  is  fed  directly  from  a low-voltage  sup- 
ply of  the  proper  potential  through  a choke  L. 
A conventional  filter  choke  having  an  induc- 
tance from  10  to  20  henries  will  be  satisfac- 
tory for  L. 

To  afford  protection  of  the  tube  when  plate 
voltage  is  not  applied  but  screen  voltage  is 
supplied  from  the  exciter  power  supply,  when 
using  the  arrangement  of  figure  9D,  a resistor 
of  3000  to  10,  000  ohms  can  be  connected  in 
series  with  the  choke  L.  In  this  case  the  screen 
supply  voltage  should  be  at  least  1}4  times  as 


HANDBOOK 


Cathode  Modulation  297 


much  as  is  required  tor  actual  screen  voltage, 
and  the  value  of  resistor  is  chosen  such  that 
with  normal  screen  current  the  drop  through 
the  resistor  and  choke  will  be  such  that  nor- 
mal screen  voltage  will  be  applied  to  the  tube. 
When  the  plate  voltage  is  removed  the  screen 
current  will  increase  greatly  and  the  drop 
through  resistor  R will  increase  to  such  a 
value  that  the  screen  voltage  will  be  lowered 
to  the  point  where  the  screen  dissipation  on 
the  tube  will  not  be  exceeded.  However,  the 
supply  voltage  and  value  of  resistor  R must 
be  chosen  carefully  so  that  the  maximum  rated 
screen  dissipation  cannot  be  exceeded.  The 
maximum  possible  screen  dissipation  using 
this  arrangement  is  equal  to:  W = EV4R  where 
E is  the  screen  supply  voltage  and  R is  the 
combined  resistance  of  the  resistor  in  figure 
9D  and  the  d-c  resistance  of  the  choke  L.  It 
is  wise,  when  using  this  arrangement  to  check, 
using  the  above  formula,  to  see  that  the  value 
of  W obtained  is  less  than  the  maximum  rated 
screen  dissipation  of  the  tube  or  tubes  used 
in  the  modulated  stage.  This  same  system  can 
of  course  also  be  used  in  figuring  the  screen 
supply  circuit  of  a pentode  or  tetrode  ampli- 
fier stage  where  modulation  is  not  to  be 
applied. 

The  modulation  transformer  for  plate-and- 
screen-modulation,  when  utilizing  a dropping 
resistor  as  shown  in  figure  9A,  is  similar  to 
the  type  of  transformer  used  for  any  plate 
modulated  phone.  The  combined  screen  and 
plate  current  is  divided  into  the  plate  voltage 
in  order  to  obtain  the  Class  C amplifier  load 
impedance.  The  peak  audio  power  required  to 
obtain  100  per  cent  modulation  is  equal  to  the 
d-c  power  input  to  the  screen,  screen  resistor, 
and  plate  of  the  modulated  t-f  stage. 

15-5  Cathode  Modulation 

Cathode  modulation  offers  a workable  com- 
promise between  the  good  plate  efficiency  but 
expensive  modulator  of  high-level  plate  modu- 
lation, and  the  poor  plate  efficiency  but  in- 
expensive modulator  of  grid  modulation.  Cathode 
modulation  consists  essentially  of  an  ad- 
mixture of  the  two. 

The  efficiency  of  the  average  well-designed 
plate-modulated  transmitter  is  in  the  vicinity 
of  75  to  80  per  cent,  with  a compromise  per- 
haps at  77.5  per  cenr.  On  the  other  hand,  the 
efficiency  of  a good  grid-modulated  transmitter 
may  run  from  28  to  maybe  40  per  cent,  with 
the  average  falling  at  about  34  pet  cent.  Now 
since  cathode  modulation  consists  of  simul- 
taneous grid  and  plate  modulation,  in  phase 
with  each  other,  we  can  theoretically  obtain 
any  efficiency  from  about  34  to  77.5  per  cent 
from  out  cathode-modulated  stage,  depending 


upon  the  relative  percentages  of  grid  and 
plate  modulation. 

Since  the  system  is  a compromise  between 
the  two  fundamental  modulation  arrangements, 
a value  of  efficiency  approximately  half  way 
between  the  two  would  seem  to  be  the  best 
compromise.  Experience  has  proved  this  to  be 
the  case.  A compromise  efficiency  of  about 
56.5  per  cent,  roughly  half  way  between  the 
two  limits,  has  proved  to  be  optimum.  Cal- 
culation has  shown  that  this  value  of  effi- 
ciency can  be  obtained  from  a cathode-modu- 
lated amplifier  when  the  audio-frequency  modu- 
lating power  is  approximately  20  per  cent  of 
the  d-c  input  to  the  cathode-modulated  stage. 

An  Economical  Series  cathode  modulation  is 
Series  Cathode  ideally  suited  as  an  economi- 
Modulotor  cal  modulating  arrangement 

for  a high-power  triode  c-w 
transmitter.  The  modulator  can  be  constructed 
quite  compactly  and  for  a minimum  component 
cost  since  no  power  supply  is  required  for  it. 
When  it  is  desired  to  change  over  from  c-w  to 
’phone,  it  is  only  necessary  to  cut  the  series 
modulator  inro  the  cathode  return  circuit  of  the 
c-w  amplifier  stage.  The  plate  voltage  for  the 
modulator  tubes  and  for  the  speech  amplifier 
is  taken  from  the  cathode  voltage  drop  of  the 
modulated  stage  across  the  modulator  unit. 

Figure  10  shows  the  circuit  of  such  a modu- 
lator, designed  to  cathode  modulate  a Class  C 
amplifier  using  push-pull  810  tubes,  running 
at  a supply  voltage  of  2500,  and  with  a plate 
input  of  660  watts.  The  modulated  stage  runs 
at  about  50%  efficiency,  giving  a power  output 
of  neatly  350  watts,  fully  modulated.  The  volt- 
age drop  across  the  cathode  modulator  is  400 
volts,  allowing  a net  plate  to  cathode  voltage 
of  2100  volts  on  the  final  amplifier.  The  plate 
current  of  the  810’ s should  be  about  330  ma., 
and  the  grid  current  should  be  approximately 
40  ma.,  making  the  total  cathode  current  of  the 
modulated  stage  370  ma.  Four  parallel  6L6 
modulator  tubes  can  pass  this  amount  of  plate 
current  without  difficulty.  It  must  be  remem- 
bered that  the  voltage  drop  across  the  cathode 
modulator  is  also  the  cathode  bias  of  the  modu- 
lated stage.  In  most  cases,  no  extra  grid  bias 
is  necessary.  If  a bias  supply  is  used  for  c-w 
operation,  it  may  be  removed  for  cathode  modu- 
lation, as  shown  in  figure  11.  With  low-mu 
triodes,  some  extra  grid  bias  (over  and  above 
that  amount  supplied  by  the  cathode  modulator) 
may  be  needed  to  achieve  proper  linearity  of 
the  modulated  stage.  In  any  case,  proper  oper- 
ation of  a cathode  modulated  stage  should  be 
determined  by  examining  the  modulated  output 
waveform  of  the  stage  on  an  oscilloscope. 

Excitation  The  r-f  driver  for  a cathode-mod- 
ulated stage  should  have  about 


THE  RADIO 


298  Amplitude  Modulation 


TO  CATHODE 

MODULATED 

STAGE 


Figure  10 

SERIES  CATHODE  MODULATOR  FOR  A HIGH-POWERED  TRIODE  R-F 

AMPLIFIER 


the  same  power  output  capabilities  as  would 
be  required  to  drive  a c-w  amplifier  to  the  same 
input  as  it  is  desired  to  drive  the  cathode- 
modulated  stage.  However,  some  form  of  exci- 
tation control  should  be  available  since  the 
amount  of  excitation  power  has  a direct  bearing 
on  the  linearity  of  a cathode-modulated  am- 
plifier stage.  If  link  coupling  is  used  between 
the  driver  and  the  modulated  stage,  variation 
in  the  amount  of  link  coupling  will  afford 
ainple  excitation  variation.  If  much  less  than 
40%  plate  modulation  is  employed,  the  stage 
begins  to  resemble  a grid-bias  modulated 
stage,  and  the  necessity  for  good  r-f  regula- 
tion will  apply. 

Cathode  Modulation  Cathode  modulation  has 

of  Tetrodes  not  proved  too  satisfac- 

tory for  use  with  beam 
tetrode  tubes.  This  is  a result  of  the  small 
excitation  and  grid  swing  requirements  for 
such  tubes,  plus  the  fact  that  some  means  for 
holding  the  screen  voltage  at  the  potential  of 
the  cathode  as  fat  as  audio  is  concerned  is 
usually  necessary.  Because  of  these  factors, 
cathode  modulation  is  not  recommended  for 
use  with  tetrode  r-f  amplifiers. 

15-6  The  Doherty  and  the 
T erman-Woodyord 
Modulated  Amplifiers 

These  two  amplifiers  will  be  described  to- 
gether since  they  operate  upon  very  similar 
principles.  Figure  12  shows  a greatly  simpli- 
fied schematic  diagram  of  the  operation  of  both 
types.  Both  systems  operate  by  virtue  of  a car- 
rier tube  (V,  in  both  figures  12  and  13)  which 


supplies  the  unmodulated  carrier,  and  whose 
output  is  reduced  to  supply  negative  peaks, 
and  a peak  tube  (Vj)  whose  function  is  to 
supply  approximately  half  the  positive  peak 
of  the  modulation  cycle  and  whose  additional 
function  is  to  lower  the  load  impedance  on  the 
carrier  tube  so  that  it  will  be  able  to  supply 
the  other  half  of  the  positive  peak  of  the  modu- 
lation cycle. 

The  peak  tube  is  enabled  to  increase  the 
output  of  the  carrier  tube  by  virtue  of  an  im- 
pedance inverting  line  between  the  plate  cir- 
cuits of  the  two  tubes.  This  line  is  designed 
to  have  a characteristic  impedance  of  one-half 
the  value  of  load  into  which  the  carrier  tube 
operates  under  the  carrier  conditions.  Then  a 
load  of  one-half  the  characteristic  impedance 
of  the  quarter-wave  line  is  coupled  into  the 
output.  By  experience  with  quarter-wave  lines 
in  antenna-matching  circuits  we  know  that 
such  a line  will  vary  the  impedance  at  one 
end  of  the  line  in  such  a manner  that  the  geo- 
metric mean  between  the  two  terminal  imped- 
ances will  be  equal  to  the  characteristic  im- 
pedance of  the  line.  Thus,  if  we  have  a value 
of  load  of  one-halj  the  characteristic  imped- 
ance of  the  line  at  one  end,  the  other  end  of 
the  line  will  present  a value  of  twice  the  char- 
acteristic impedance  of  the  lines  to  the  car- 
rier tube  Vi. 

This  is  the  situation  that  exists  under  the 
carrier  conditions  when  the  peak  tube  merely 
floats  across  the  load  end  of  the  line  and  con- 
tributes no  power.  Then  as  a positive  peak  of 
modulation  comes  along,  the  peak  tube  starts 
to  contribute  power  to  the  load  until  at  the 
peak  of  the  modulation  cycle  it  is  contributing 
enough  power  so  that  the  impedance  at  the 
load  end  of  the  line  is  equal  to  R,  instead  of 


HANDBOOK 


Doherty  Amplifier  299 


R.F.  AMPLIFIER 


CATHODE  MODULATOR  INSTALLATION 
SHOWING  PHONE-C.W.  TRANSFER  SWITCH 


the  R/2  that  is  presented  under  the  carrier 
conditions.  This  is  true  because  at  a positive 
modulation  peak  (since  it  is  delivering  full 
power)  the  peak  tube  subtracts  a negative 
resistance  of  R/2  from  the  load  end  of  the 
line. 

Now,  since  under  the  peak  condition  of  modu- 
lation the  load  end  of  the  line  is  terminated 
in  R ohms  instead  of  R/2,  the  impedance  at 
the  carrier-tube  will  be  reduced  from  2R  ohms 
to  R ohms.  This  again  is  due  to  the  impedance 
inverting  action  of  the  line.  Since  the  load  re- 
sistance on  the  carrier  tube  has  been  reduced 
to  half  the  carrier  value,  its  output  at  the  peak 
of  the  modulation  cycle  will  be  doubled.  Thus 
we  have  the  pecessary  condition  for  a 100 
per  cent  modulation  peak;  the  amplifier  will 
deliver  four  times  as  much  power  as  it  does 
under  the  carrier  conditions. 

On  negative  modulation  peaks  the  peak  tube 
does  not  contribute;  the  output  of  the  carrier 
tube  is  reduced  until  on  a 100  per  cent  nega- 
tive peak  its  output  is  zero. 

The  Electrical  While  an  electrical  quarter- 

Quarter-Wave  wave  line  (consisting  of  a pi 
Line  network  with  the  inductance 

and  capacitance  units  having 
a reactance  equal  to  the  characteristic  imped- 
ance of  the  line)  does  have  the  desired  im- 
pedance-inverting effect,  it  also  has  the  un- 
desirable effect  of  introducing  a 90°  phase 
shift  across  such  a line.  If  the  shunt  elements 

are  capacitances,  the  phase  shift  across  the 
line  lags  by  90°;  if  they  ate  inductances,  the 
phase  shift  leads  by  90°.  Since  there  is  an  un- 


ELECTRICAL  A/* 


DIAGRAMMATIC  REPRESENTATION  OF 
THE  DOHERTY  LINEAR 


desirable  phase  shift  of  90°  between  the  plate 
circuits  of  the  carrier  and  peak  tubes,  an  equal 
and  opposite  phase  shift  must  be  introduced  in 
the  exciting  voltage  to  the  grid  circuits  of  the 
two  tubes  so  that  the  resultant  output  in  the 
plate  circuit  will  be  in  phase.  This  additional 
phase  shift  has  been  indicated  in  figure  12  and 
a method  of  obtaining  it  has  been  shown  in 
figure  13. 

Comparison  Between  The  difference  between 
Linear  ond  the  Doherty  linear  am- 

Grid  Modulator  plifier  and  the  Terman- 

Woodyard  grid-modulated 
amplifier  is  the  same  as  the  difference  between 
any  linear  and  grid-modulated  stages.  Modulated 
r.f.is  applied  to  the  grid  circuit  of  the  Doherty 
linear  amplifier  with  the  carrier  tube  biased  to 
cutoff  and  the  peak  tube  biased  to  the  point 
where  it  draws  substantially  zero  plate  current 
at  the  carrier  condition. 

In  the  Terman- Woody ard  grid-modulated  am- 
plifier the  carrier  tube  runs  Class  C with  com- 
paratively high  bias  and  high  plate  efficiency, 
while  the  peak  tube  again  is  biased  so  that  it 
draws  almost  no  plate  current.  Unmodulated 
r.f.  is  applied  to  the  grid  circuits  of  the  two 
tubes  and  the  modulating  voltage  is  inserted 
in  series  with  the  fixed  bias  voitages.  From 
one-half  to  two-thirds  as  much  audio  voltage 
is  required  at  the  grid  of  the  peak  tube  as  is 
required  at  the  grid  of  the  carrier  tube. 

Operating  The  resting  carrier  efficiency  of 

Efficiencies  the  grid-modulated  amplifier  may 

run  as  high  as  is  obtainable  in 
any  Class  C stage,  80  per  cent  or  better.  The 
resting  carrier  efficiency  of  the  linear  will  be 
about  as  good  as  is  obtainable  in  any  Class 
B amplifier,  60  to  70  per  cent.  The  overall 
efficiency  of  the  bias-modulated  amplifier  at 
100  pet  cent  modulation  will  run  about  75  per 
cent;  of  the  linear,  about  60  per  cent. 

In  figure  13  the  plate  tank  circuits  are  de- 
tuned enough  to  give  an  effect  equivalent  to 
the  shunt  elements  of  the  quarter-wave  "line” 
of  figure  12.  At  resonance,  the  coils  L,  and 
L2  in  the  grid  circuits  of  the  two  tubes  have 


300  Amplitude  Modulation 


THE  RADIO 


"HIGH  EFFICIENCY"  AMPLIFIER 
The  basic  system,  comprising  a carrier** 
tube  and  a "peak**  tube  intereonnecfed  by 
lumped-constant  quarter-wave  lines,  is  the 
same  for  either  grid-bias  modulation  or  for 
use  as  a linear  amplifier  of  a modulated 
wave. 


each  an  inductive  reactance  equal  to  the  ca- 
pacitive reactance  of  the  capacitor  Q.  Thus 
we  have  the  effect  of  a pi  network  consisting 
of  shunt  inductances  and  series  capacitance. 
In  the  plate  circuit  we  want  a phase  shift  of 
the  same  magnitude  but  in  the  opposite  direc- 
tion; so  our  series  element  is  the  inductance 
L3  whose  reactance  is  equal  to  the  character- 
istic impedance  desired  of  the  network.  Then 
the  plate  tank  capacitors  of  the  two  tubes  C2 
and  C3  ate  increased  an  amount  past  reson- 
ance, so  that  they  have  a capacitive  reactance 
equal  to  the  inductive  reactance  of  the  coil  L3. 
It  is  quite  important  that  there  be  no  coupling 
between  the  inductors. 

Although  both  these  types  of  amplifiers  are 
highly  efficient  and  requite  no  high-level  audio 
equipment,  they  are  difficult  to  adjust— parti- 
cularly so  on  the  higher  frequencies— and  it 
would  be  an  extremely  difficult  problem  to  de- 
sign a multiband  transmitter  employing  the 
circuit.  However,  the  grid-bias  modulation  sys- 
tem has  advantages  for  the  high-power  trans- 
mitter which  will  be  operated  on  a single  fre- 
quency band. 

Other  High-Efficiency  Many  other  high-efficien- 
Modulation  Systems  cy  modulation  systems 
have  been  described  since 
about  1936.  The  majority  of  these,  however 
have  received  little  application  either  by  com- 
mercial interests  or  by  amateurs.  In  most  cases 
the  circuits  are  difficult  to  adjust,  or  they  have 


Other  undesirable  features  which  make  their 

use  impracticable  alongside  the  more  conven- 
tional modulation  systems.  Nearly  all  these 
circuits  have  been  published  in  the  I.R.E. 
Proceedings  and  the  interested  reader  can  re- 
fer to  them  in  back  copies  of  that  journal. 


15.7  Speech  Clipping 

Speech  waveforms  are  characterized  by  fre- 
quently recurring  high-intensity  peaks  of  very 
short  duration.  These  peaks  will  cause  over- 
modulation if  the  average  level  of  modulation 
on  loud  syllables  exceeds  approximately  30 
per  cent.  Careful  checking  into  the  nature  of 
speech  sounds  has  revealed  that  these  high- 
intensity  peaks  are  due  primarily  to  the  vowel 
sounds.  Further  research  has  revealed  that  the 
vowel  sounds  add  little  to  intelligibility,  the 
major  contribution  to  intelligibility  coming 
from  the  consonant  sounds  such  as  v,  b,  k,  s, 
t,  and  1.  Measurements  have  shown  that  the 
power  contained  in  these  consonant  sounds 
may  be  down  30  db  or  more  from  the  energy  in 
the  vowel  sounds  in  the  same  speech  passage. 
Obviously,  then,  if  we  can  increase  the  rel- 
ative energy  content  of  the  consonant  sounds 
with  respect  to  the  vowel  sounds  it  will  be 
possible  to  understand  a signal  modulated  with 
such  a waveform  in  the  presence  of  a much 
higher  level  of  background  noise  and  inter- 
ference. Experiment  has  shown  that  it  is  pos- 
sible to  accomplish  this  desirable  result  sim- 
ply by  cutting  off  or  clipping  the  high-intensity 
peaks  and  thus  building  up  in  a relative  man- 
ner the  effective  level  of  the  weaker  sounds. 

Such  clipping  theoretically  can  be  accom- 
plished simply  by  increasing  the  gain  of  the 
speech  amplifier  until  the  average  level  of 
modulation  on  loud  syllables  approaches  90 
per  cent.  This  is  equivalent  to  increasing  the 
speech  power  of  the  consonant  sounds  by  about 
10  times  or,  conversely,  we  can  say  that  10  db 
of  clipping  has  been  applied  to  the  voice  wave. 
However,  the  clipping  when  accomplished  in 
this  manner  will  produce  higher  order  side- 
bands known  as  "splatter,”  and  the  transmitted 
signal  would  occupy  a relatively  tremendous 
slice  of  spectrum.  So  another  method  of  accom- 
plishing the  desirable  effects  of  clipping  must 
be  employed. 

A considerable  reduction  in  the  amount  of 
splatter  caused  by  a moderate  increase  in  the 
gain  of  the  speech  amplifier  can  be  obtained 
by  poling  the  signal  from  the  speech  amplifier 
to  the  transmitter  such  that  the  high-intensity 
peaks  occur  on  upward  or  positive  modulation. 
Overloading  on  positive  modulation  peaks  pro- 
duces less  splatter  than  the  negative-peak 
clipping  which  occurs  with  overloading  on  the 


HANDBOOK 


Speech  Clipping  301 


Figure  14 

SPEECH-WAVEFORM  AMPLITUDE 
MODULATION 

Showing  the  effect  of  using  the  prop- 
er polarity  of  a speech  wave  for 
modulating  a tronsmitter.  fA)  shows 
effect  of  proper  speech  polarity 
on  o transmitter  having  an  upward 
modulation  capability  of  greater 
than  700  per  cent.  (B)  shows  the 
effect  of  using  proper  speech  polar- 
ity on  a transmitter  hoving  on  op- 
ward  modulation  capability  of  only 
700  per  cent.  Both  these  conditions 
will  give  a clean  signal  without 
objectionable  sp/offer.  (C)  shows 
the  effect  of  the  use  of  improper 
speech  polarity.  This  condition  will 
cause  serious  splatter  due  to  nega- 
tive-peak clipping  in  the  modulated- 
ampllfier  stage. 


jg0_%_TOS^M^ULAT[O2i 


negative  peaks  of  modulation.  This  aspect  of 
the  problem  has  been  discussed  in  more  detail 
in  the  section  on  Speech  Waveform  Dissymmetry 
earlier  in  this  chapter.  The  effect  of  feeding 
the  proper  speech  polarity  from  the  speech  am- 
plifier is  shown  in  figure  14. 

A much  more  desirable  and  effective  method 
of  obtaining  speech  clipping  is  actually  to  em- 
ploy a clipper  circuit  in  the  earlier  stages  of 
the  speech  amplifier,  and  then  to  filter  out  the 
objectionable  distortion  components  by  means 
of  a sharp  low-pass  filter  having  a cut-off  fre- 
quency of  approximately  3000  cycles.  Tests  on 
clipper-lilter  speech  systems  have  shown  that 
6 db  of  clipping  on  voice  is  just  noticeable, 
12  db  of  clipping  is  quite  acceptable,  and 
values  of  clipping  from  20  to  25  db  are  toler- 
able under  such  conditions  that  a high  degree 
of  clipping  is  necessary  to  get  through  heavy 
QRM  or  QRN.  A signal  with  12  db  of  clipping 
doesn’t  sound  quite  natural  but  it  is  not  un- 
pleasant to  listen  to  and  is  much  more  read- 
able than  an  undipped  signal  in  the  presence 
of  strong  interference. 

The  use  of  a clipper-filter  in  the  speech  am- 

plifier,  to  be  completely  effective,  requires 

that  phase  shift  between  the  clipper-filter 
stage  and  the  final  modulated  amplifier  be  kept 


to  a minimum.  However,  if  there  is  phase  shift 
after  the  clipper-filter  the  system  does  not 
completely  break  down.  The  presence  of  phase 
shift  merely  requires  that  the  audio  gain  fol- 
lowing the  clipper-filter  be  reduced  to  the  point 
where  the  cant  applied  to  the  clipped  speech 
waves  still  cannot  cause  overmodulation.  This 
effect  is  illustrated  in  figures  15  and  16. 

The  cant  appearing  on  the  tops  of  the  square 
waves  leaving  the  clipper-filter  centers  about 
the  clipping  level.  Hence,  as  the  frequency 
being  passed  through  the  system  is  lowered, 
the  amount  by  which  the  peak  of  the  canted 
wave  exceeds  the  clipping  level  is  increased. 

Phase  Shift  In  a normal  transmitter  having  a 
Correction  moderate  amount  of  phase  shift 
the  cant  applied  to  the  tops  of 
the  waves  will  cause  overmodulation  on  fre- 
quencies below  those  for  which  the  gain  fol- 
lowing the  clipper-filter  has  been  adjusted  un- 
less remedial  steps  have  been  taken.  The  fol- 
lowing steps  are  advised: 

(1)  Introduce  bass  suppression  into  the  speech 
amplifier  ahead  of  the  clipper-filter. 

(2)  Improve  the  low-frequency  response  char- 
acteristic insofar  as  it  is  possible  in  the 


3 02  Amplitude  Modulation 


THE  RADIO 


CLIPPED  AND  FILTERED  SPEECH  WAVE 


too  ^PMITIVE  MOOULATI^ 
70%  POSlTjye  MODULi^lM 


MODULATED  WAVE  AFTER  UNDERGOING  PHASE  SHIFT 


Figure  15 

ACTION  OF  A CLIPPER-FILTER 
ON  A SPEECH  WAVE 
7/*e  drawing  (A)  shows  the  incom- 
ing spaach  wave  baforo  it  roaches 
the  dipper  stage.  (B)  shows  the 
output  of  the  clipper-filter,  illus- 
trtting  the  manner  in  which  the 
peaks  are  clipped  and  then  the 
sharp  edges  of  the  clipped  wave 
removed  by  the  filter.  (C)  shows 
the  e^ect  of  phase  shift  in  the 
stages  following  the  cl ipper-f liter. 
(C)  also  shows  the  manner  in  which 
the  transmitter  may  be  adjusted  for 
100  per  cent  modulation  of'  tho 
**  eantod'*  poaks  of  tho  wavo,  tho 
sloping  top  of  tho  wave  roaching 
about  70  pot  cent  modulation. 


stages  following  the  clippet-filtet.  Feed- 
ing the  plate  current  to  the  final  amplifier 
through  a choke  rather  than  through  the 
secondary  of  the  modulation  transformer 
will  help  materially. 

Even  with  the  normal  amount  of  improvement 
which  can  be  attained  through  the  steps  men- 
tioned above  there  will  still  be  an  amount  of 
wave  cant  which  must  be  compensated  in  some 
manner.  This  compensation  can  be  done  in 
either  of  two  ways.  The  first  and  simpler  way 
is  as  follows: 

(1)  Adjust  the  speech  gain  ahead  of  the  clip- 
per-filter until  with  normal  talking  into 
the  microphone  the  distortion  being  intro- 
duced by  the  clipper-filter  circuit  is  quite 
apparent  but  not  objectionable.  This  amount 
of  distortion  will  be  apparent  to  the  normal 
listener  when  10  to  15  db  of  clipping  is 
taking  place. 

( 2)  Tune  a selective  communications  receiver 
about  15  kc.  to  one  side  or  the  other  of  the 
frequency  being  transmitted.  Use  a short 
antenna  or  no  antenna  at  all  on  the  re- 
ceiver so  that  the  transmitter  is  not  block- 
ing the  receiver. 


(3)  Again  with  the  normal  talking  into  the 
microphone  adjust  the  gain  following  the 
clipper-filter  to  the  point  where  the  side- 
band splatter  is  being  heard,  and  then 
slightly  back  off  the  gain  after  the  clip- 
pet-filter  until  the  splatter  disappears. 

If  the  phase  shift  in  the  transmitter  or  mod- 
ulator is  not  excessive  the  adjustment  proced- 
ure given  above  will  allow  a clean  signal  to  be 
radiated  regardless  of  any  reasonable  voice 
level  being  fed  into  the  microphone. 

If  a cathode-ray  oscilloscope  is  available 
the  modulated  envelope  of  the  transmitter 
should  be  checked  with  30  to  70  cycle  saw- 
tooth waves  on  the  horizontal  axis.  If  the  upper 
half  of  the  envelope  appears  in  general  the 
same  as  the  drawing  of  figure  15C,  all  is  well 
and  phase-shift  is  not  excessive.  However,  if 
much  more  slope  appears  on  the  tops  of  the 
waves  than  is  illustrated  in  this  figure,  it  will 
be  well  to  apply  the  second  step  in  compen- 
sation in  order  to  insure  that  sideband  splatter 
cannot  take  place  and  to  afford  a still  higher 
average  percentage  of  modulation.  This  second 
step  consists  of  the  addition  of  a high-level 
splatter  suppressor  such  as  is  illustrated  in 
figure  17. 


HANDBOOK 


Splatter  Suppression  303 


3000  O,  WAVE 


Figure  16 

ILLUSTRATING  THE  EFFECT  OF  PHASE 
SHIFT  AND  FILTERED  WAVES  OF  DIF- 
FERENT FREQUENCY 
Sketch  (A)  shows  the  effect  of  a clipper  and 
a filter  having  a cutoff  of  about  3500  cycles 
on  a wave  of  3000  cycles.  Note  that  no  har- 
monics are  present  in  the  wave  sa  that  phase 
shift  foilawing  the  clipper-filter  will  have 
no  significant  effect  on  the  shape  of  the 
wave.  (B)  and  (C)  show  the  effect  of  phase 
shift  on  waves  well  below  the  cutoff  frequen- 
cy of  the  filter.  Note  that  the  "cant"  placed 
upon  the  top  of  the  wave  causes  the  peak 
value  to  rise  higher  and  higher  above  the 
dipping  level  as  the  frequency  is  lowered. 
It  is  for  this  reason  that  bass  suppression 
before  the  clipper  stage  is  desirable.  Im- 
proved low-frequency  response  following  the 
dipper-filter  will  reduce  the  phase  shift  and 
therefore  the  canting  of  the  wave  at  the  lower 
voice  frequencies. 


The  use  of  a high-level  splatter  suppressor 
after  a clipper-filter  system  will  afford  the  re- 
sult shown  in  figure  18  since  such  a device 
will  not  permit  the  negative-peak  clipping 
which  the  wave  cant  caused  by  audio-system 
phase  shift  can  produce.  The  high-level  splat- 
ter suppressor  operates  by  virtue  of  the  fact 
that  it  will  not  permit  the  plate  voltage  on  the 
modulated  amplifier  to  go  completely  to  zero 
regardless  of  the  incoming  signal  amplitude. 


Figure  17 

HIGH-LEVEL  SPLATTER  SUPPRESSOR 

This  circuit  is  effective  in  reducing  splatter 
caused  by  negative-peak  clipping  in  the  mod- 
ulated amplifier  stage.  The  use  of  a two- 
section  filter  as  shown  is  recommended,  al- 
though either  a single  m-derived  or  a con- 
stant-k  section  may  be  used  for  greater  econ- 
omy. Suitable  chokes,  along  with  recom- 
mended capacitor  values,  are  available  from 
several  manufacturers. 


Hence  negative-peak  clipping  with  its  attend- 
ant splatter  cannot  take  place.  Such  a device 
can,  of  course,  also  be  used  in  a transmitter 
which  does  not  incorporate  a clipper-filter  sys- 
tem. However,  the  full  increase  in  average 
modulation  level  without  serious  distortion, 
afforded  by  the  clipper-filter  system,  will  not 
be  obtained. 

A word  of  caution  should  be  noted  at  this 
time  in  the  case  of  tetrode  final  modulated 
amplifier  stages  which  afford  screen  voltage 
modulation  by  virtue  of  a tap  or  a separate 
winding  on  the  modulation  transformer  such 
as  is  shown  in  figure  9C  of  this  chapter.  If 
such  a system  of  modulation  is  in  use,  the 
high-level  splatter  suppressor  shown  in  figure 
17  will  not  operate  satisfactorily  since  nega- 
tive-peak clipping  in  the  stage  can  take  place 
when  the  screen  voltage  goes  too  low. 

Clipper  Circuits  Two  effective  low-level  clip- 
pet-filter  circuits  are  shown 
in  figures  19  and  20.  The  circuit  of  figure  19 
employs  a 6J6  double  triode  as  a clipper,  each 
half  of  the  6J6  clipping  one  side  of  the  im- 
pressed waveform.  The  optimum  level  at  which 
the  clipping  operation  begins  is  set  by  the 
value  of  the  cathode  resistor.  A maximum  of 
12  to  14  db  of  clipping  may  be  used  with  this 
circuit,  which  means  that  an  extra  12  to  14  db 
of  speech  gain  must  precede  the  clipper.  For 
a peak  output  of  8 volts  from  the  clipper-filter, 
a peak  audio  signal  of  about  40  volts  must  be 
impressed  upon  the  dipper  input  circuit.  The 
6C4  speech  amplifier  stage  must  therefore  be 
considered  as  a part  of  the  clipper  circuit  as 


304  Amplitude  Modulation 


THE  RADIO 


Figure  18 

ACTION  OF  HIGH-LEVEL 
SPLATTER  SUPPRESSOR 

A high-level  splatter  suppressor 
may  be  used  in  a transmitter  with- 
out a clipper-filter  to  reduce  nega- 
tive-peak clipping,  or  such  a unit 
may  be  used  following  a clipper- 
filter  to  allow  a higher  overage 
modulation  level  by  eliminating  the 
negative-peak  clipping  which  the 
wave-cant  caused  by  phase  shift 
might  produce. 


it  compensates  for  the  12  to  14  db  loss  of  gain 
incurred  in  the  clipping  process.  A simple  low- 
pass  filter  made  up  of  a 20  henry  a.c.  - d.c 
replacement  type  filter  choke  and  two  mica 
condensers  follows  the  6J6  clipper.  This  fil- 
ter is  designed  for  a cutoff  frequency  of  about 
3500  cycles  when  operating  into  a load  im- 
pedance of  ^ megohm.  The  output  level  of  8 
volts  peak  is  ample  to  drive  a triode  speech 
amplifier  stage,  such  as  a 6C4  or  6J5. 

A 6AL5  double  diode  series  clipper  is  em- 
ployed in  the  circuit  of  figure  20,  and  a com- 
mercially made  low-pass  filter  is  used  to  give 
somewhat  better  high  frequency  cutoff  char- 
acteristics. A double  triode  is  employed  as  a 
speech  amplifier  ahead  of  the  clipper  circuit. 
The  actual  performance  of  either  circuit  is 
about  the  same. 

To  eliminate  higher  order  products  that  may 
be  generated  in  the  stages  following  the  clip- 
per-filter, it  is  wise  to  follow  the  modulator 
with  a high-level  filter,  as  shown  in  figure  21. 

Clipper  Adiustment  These  clipper  circuits 
have  two  adjustments: 
Adjust  Gain  and  Adjust  Clipping.  The  Adj. 
Gain  control  determines  the  modulation  level 
of  the  transmitter.  This  control  should  be  set 


so  that  over-modulation  of  the  transmitter  is 
impossible,  regardless  of  the  amount  of  clip- 
ping used.  Once  the  Adj.  Gain  control  has 
been  roughly  set,  the  Adj.  Clip,  control  may 
be  used  to  set  the  modulation  level  to  any  per- 
centage below  100%.  As  the  modulation  level 
is  decreased,  more  and  more  clipping  is  intro- 
duced into  the  circuit,  until  a full  12  db  of 
clipping  is  used.  This  means  that  the  Adj. 
Gain  control  may  be  advanced  some  12  db  past 
the  point  where  the  clipping  action  started. 
Clipping  action  should  start  at  85%  to  90% 
modulation  when  a sine  wave  is  used  for  cir- 
cuit adjustment  purposes. 

High-Level  Even  though  we  may  have  cut  off 
Filters  all  frequencies  above  3000  or  3500 
cycles  through  the  use  of  a filter 
system  such  as  is  shown  in  the  circuits  of  fig- 
ures 19  and  20,  higher  frequencies  may  again 
be  introduced  into  the  modulated  wave  by 
distortion  in  stages  following  the  speech  am- 
plifier. Harmonics  of  the  incoming  audio  fre- 
quencies may  be  generated  in  the  driver  stage 
for  the  modulator;  they  may  be  generated  in  the 
plate  circuit  of  the  modulator;  or  they  may  be 
generated  by  non-linearity  in  the  modulated 
amplifier  itself. 


20  H 
(STANCOR  C-1515) 


ADJUST  GAIN 


Figure  19 

CLIPPER  FILTER  USING  6J6  DOUBLE  TRIODE  STAGE 


HANDBOOK 


Splatter  Suppression  305 


Regardless  of  the  point  in  the  system  fol- 
lowing the  speech  amplifier  where  the  high 
audio  frequencies  may  be  generated,  these  fre- 
quencies can  still  cause  a broad  signal  to  be 
transmitted  even  though  all  frequencies  above 
3000  or  3500  cycles  have  been  cut  off  in  the 
speech  amplifiet.  The  effects  of  distortion  in 
the  audio  system  following  the  speech  ampli- 
fiet can  be  eliminated  quite  effectively  through 
the  use  of  a post-modulator  filter.  Such  a filter 
must  be  used  between  the  modulator  plate  cir- 
cuit and  the  r-f  amplifier  which  is  being  mod- 
ulated. 

This  filtet  may  take  three  general  forms  in 
a normal  case  of  a Class  C amplifier  plate  mod- 
ulated by  a Class  B modulator.  The  best  meth- 
od is  to  use  a high  level  low-pass  filter  as 


CLASS  C AMPLIFIER 


B+  MOD.  6+  R.F. 


Figure  21 

ADDITIONAL  HIGH-LEVEL  LOW-PASS  FIL- 
TER TO  FOLLOW  MODULATOR  WHEN  A 
LOW-LEVEL  CLIPPER  FILTER  IS  USED 
Suitable  choke,  along  with  recommenc/ed  ca- 
pacitor values,  is  available  from  several 
manufacturers. 


shown  in  figure  21  and  discussed  previously. 
Another  method  which  will  give  excellent  re- 
sults in  some  cases  and  poor  results  in  others, 
dependent  upon  the  characteristics  of  the  mod- 
ulation transformer,  is  to  **build  out’*  the  mod- 
ulation transformer  into  a filter  section.  This 
is  accomplished  as  shown  in  figure  22  by  plac- 
ing mica  capacitors  of  the  correct  value  across 
the  primary  and  secondary  of  the  modulation 
transformer.  The  proper  values  for  the  capaci- 


CLASS  C STAGE 


Figure  22 

"BUILDING-OUT"  THE  MODULATION 
TRANSFORMER 

This  expedient  utilizes  the  leakage  reacf- 
once  of  the  modulation  transformer  in  con- 
junction with  the  capacitors  shown  to  make 
up  a single-section  low-pass  filter.  In  order 
to  determine  exact  values  for  C;  and  C2  plus 
C3,  it  is  necessary  to  use  a measurement 
setup  such  as  is  shown  in  figure  23.  How- 
ever, experiment  has  shown  in  the  case  of  a 
number  of  commercially  available  modulation 
transformers  that  a value  for  Cj  of  0,002-/J-fd, 

and  C2  plus  C3  of  O.OOd-fifd.  will  give  satis- 
factory results. 


306  Amplitude  Modulation 


THE  RADIO 


Figure  23 

TEST  SETUP  FOR  BUILDING-OUT 
MODULATION  TRANSFORMER 
Through  the  use  of  a test  setup  such  as  is 
shown  and  the  method  described  in  the  text 
it  is  possible  to  determine  the  correct  values 
for  a specified  filter  characteristic  In  the 
built-out  modulation  transformer. 


tors  Q and  C2  must,  in  the  ideal  case,  be  de- 
termined by  trial  and  error.  Experiment  with 
a number  of  modulators  has  shown,  however, 
that  if  a 0.002  fifd.  capacitor  is  used  for 
and  if  the  sum  of  C2  and  Q is  made  0.  004  pfd. 
(0.002  ;ifd.  for  C2  and  0.  002  for  C3)  the  ideal 
condition  of  cutoff  above  3000  cycles  will  be 
approached  in  most  cases  with  the  "multiple- 
match”  type  of  modulation  transformer. 

If  it  is  desired  to  determine  the  optimum 
values  of  the  capacitors  across  the  transformer 
this  can  be  determined  in  several  ways,  all  of 
which  require  the  use  of  a calibrated  audio 
oscillator.  One  way  is  diagrammed  in  figure  23. 
The  series  resistors  Rj  and  R2  should  each  be 
equal  to  14  the  value  of  the  recommended  plate- 
to-plate  load  resistance  for  the  Class  B modu- 
lator tubes.  Resistor  R3  should  be  equal  to  the 
value  of  load  resistance  which  the  Class  C 
modulated  stage  will  present  to  the  modulator. 
The  meter  V can  be  any  type  of  a-c  voltmeter. 
The  indicating  instrument  on  the  secondary  of 
the  transformer  can  be  either  a cathode-ray 
oscilloscope  or  a high-impedance  a-c  volt- 
meter of  the  vacuum-tube  or  rectifier  type. 

With  a set-up  as  shown  in  figure  23  a plot  of 
output  voltage  against  frequency  is  made,  at 
all  times  keeping  the  voltage  across  V con- 
stant, using  various  values  of  capacitance  for 
Cl  and  C2  plus  C3.  When  the  proper  values  of 
capacitance  have  been  determined  which  give 
substantially  constant  output  up  to  about  3000 
or  3500  cycles  and  decreasing  output  at  all  fre- 
quencies above,  high-voltage  mica  capacitors 
can  be  substituted  if  receiving  types  were  used 
in  the  tests  and  the  transformer  connected  to 
the  modulator  and  Class  C amplifier. 

With  the  transformer  reconnected  in  the 
transmitter  a check  of  the  modulated-wave 
output  of  the  transmitter  should  be  made  using 
an  audio  oscillator  as  signal  generator  and  an 
oscilloscope  coupled  to  the  transmitter  output. 
With  an  input  signal  amplitude  fed  to  the  speech 


Figure  24 

BASE  ATTENUATION  CHART 
Frequency  attenuation  caused  by  various 
values  of  coupling  capacitor  with  a grid  re- 
sistor of  0.5  megohm  in  the  following  stage 

(Rg  > RU 


amplifier  of  such  amplitude  that  limiting  does 
not  take  place,  a substantially  clean  sine  wave 
should  be  obtained  on  the  carrier  of  the  trans- 
mitter at  all  input  frequencies  up  to  the  cutoff 
frequency  of  the  filter  system  in  the  speech 
amplifier  and  of  the  filter  which  includes  the 
modulation  transformer.  Above  these  cutoff 
frequencies  very  little  modulation  of  the  carrier 
wave  should  be  obtained.  To  obtain  a check 
on  the  effectiveness  of  the  "built  out”  modu- 
lation transformer,  the  capacitors  across  the 
primary  and  secondary  should  be  removed  for 
the  test.  In  most  cases  a marked  deterioration 
in  the  waveform  output  of  the  modulator  will 
be  noticed  with  frequencies  in  the  voice  range 
from  500  to  1500  cycles  being  fed  into  the 
speech  amplifier. 

A filter  system  similar  to  that  shown  in  fig- 
ure 17  may  be  used  between  the  modulator 
and  the  modulated  circuit  in  a grid-modulated 
or  screen-modulated  transmitter.  Lower-voltage 
capacitors  and  low-current  chokes  may  of 
course  be  employed. 

Boss  Suppression  Most  of  the  power  repre- 
sented by  ordinary  speech 
(particularly  the  male  voice)  lies  below  1000 
cycles.  If  all  frequencies  below  400  or  500 
cycles  are  eliminated  or  substantially  atten- 
uated, there  is  a considerable  reduction  in 
power  but  insignificant  reduction  in  intelligi- 


HANDBOOK 


Bass  Suppression  307 


bility.  This  means  that  the  speech  level  may 
be  increased  considerably  without  overmodu- 
lation or  overload  of  the  audio  system.  In 
addition,  if  speech  clipping  is  used,  attenua- 
tion of  the  lower  audio  frequencies  before  the 
clipper  will  reduce  phase  shift  and  canting  of 
the  clipper  output. 

A simple  method  of  bass  suppression  is  to 
reduce  the  size  of  the  interstage  coupling  ca- 
pacitors in  a resistance  coupled  amplifier. 
Figure  24  shows  the  frequency  characteristics 
caused  by  such  a suppression  circuit.  A sec- 
ond simple  bass  suppression  circuit  is  to 
place  a small  a.c.  - d.c.  type  filter  choke  from 
grid  to  ground  in  a speech  amplifier  stage,  as 
shown  in  figure  25- 

Modulated  Amplifier  The  systems  described 
Distortion  in  the  preceding  para- 

graphs will  have  no  effect 
in  reducing  a broad  signal  caused  by  non- 
linearity in  the  modulated  amplifier.  Even 
though  the  modulating  waveform  impressed  up- 
on the  modulated  stage  may  be  distortion  free, 
if  the  modulated  amplifier  is  non-linear  dis- 
tortion will  be  generated  in  the  amplifier.  The 
only  way  in  which  this  type  of  distortion  may 
be  corrected  is  by  making  the  modulated  am- 
plifier more  linear.  Degenerative  feedback 
which  includes  the  modulated  amplifier  in  the 
loop  will  help  in  this  regard. 

Plenty  of  grid  excitation  and  high  grid  bias 
will  go  a long  way  toward  making  a plate- 
modulated  Class  C amplifier  linear,  although 
such  operating  conditions  will  make  more  diffi- 
cult the  problem  of  TVI  reduction.  If  this  still 
does  not  give  adequate  linearity,  the  preced- 
ing buffer  stage  may  be  modulated  50  per  cent 
or  so  at  the  same  time  and  in  the  same  phase 
as  the  final  amplifier.  The  use  of  a grid  leak 
to  obtain  the  majority  of  the  bias  for  a Class 
C stage  will  improve  its  linearity. 

The  linearity  of  a grid-bias  modulated  r-f 
amplifier  can  be  improved,  after  proper  ad- 
justments of  excitation,  grid  bias,  and  antenna 
coupling  have  been  made  by  modulating  the 
stage  which  excites  the  grid-modulated  ampli- 
fier. The  preceding  driver  stage  may  be  grid- 
bias  modulated  or  it  may  be  plate  modulated. 
Modulation  of  the  driver  stage  should  be  in 
the  same  phase  as  that  of  the  final  modulated 
amplifier. 

15-8  The  Bias-Shift 

Heising  Modulator 

The  simple  Class  A modulator  is  limited  to 
an  efficiency  of  about  30%,  and  the  tube  must 
dissipate  the  full  power  input  during  periods 

of  quiescence.  Class  AD  and  class  B audio 

systems  have  largely  taken  the  place  of  the 
old  Heising  modulator  because  of  this  great 


+B 


"10  HENRY"  MIDGET 

AC-DC  FIL  TER  CHOKE 
[STANCOR  C-/333  ) 


Figure  25 

USE  OF  PARALLEL  INDUCTANCE 
FOR  BASS  SUPPRESSION 


waste  of  power.  It  is  possible,  however,  to 
vary  the  operating  bias  of  the  class  A modu- 
lator in  such  a way  as  to  allow  class  A oper- 
ation only  when  an  audio  signal  is  applied  to 
the  grid  of  the  tube.  During  resting  periods, 
the  bias  can  be  shifted  to  a higher  value, 
dropping  the  resting  plate  current  and  plate 
dissipation  of  the  tube.  When  voice  waveforms 
having  low  average  power  are  employed,  the 
efficiency  of  the  system  is  comparable  to  the 
popular  class  B modulator. 

The  characteristic  curve  for  a class  A modu- 
lator is  shown  in  figure  26.  Normal  bias  is 
used,  and  the  operating  point  is  placed  in  the 
middle  of  the  linear  portion  of  theEg-Ip  curve. 
Maximum  plate  input  is  limited  by  the  plate 
dissipation  of  the  tube  under  quiescent  con- 
dition. The  bias-shift  modulator  is  biased 
close  to  plate  current  cut-off  under  no  signal 
condition  (figure  27).  Resting  plate  current 


+lp 


liRlD  INPUT 
SIGNAL 


Figure  26 

CHARACTERISTIC  GRID 
VOLTAGE-PLATE  CURRENT 
CURVE  FOR  CLASS  A 
HEISING  MODULATOR 


308 


Amplitude  Modulation 


THE  RADIO 


-I-Lp 


Figure  27 

BIAS-SHIFT  MODULATOR 
OPERATING  CHARACTERISTICS 

Modulator  is  biased  closeto  plate 
current  cut-off  under  no  signal, 
condition,  B.  Upon  appi ication 
of  audio  signal,  the  bias  of  the 
stage  is  shifted  toward  the  class 
A operating  point,  A.  Bias-shift 
voltage  is  obtained  from  audio 
signal. 


and  plate  dissipation  are  therefore  quite  low. 
Upon  application  of  an  audio  signal,  the  bias 
of  the  stage  is  shifted  toward  the  class  A 
operating  point,  preventing  the  negative  peaks 
of  the  applied  audio  voltage  from  cutting  off 
the  plate  current  of  the  tube.  As  the  audio 
voltage  increases,  the  operating  bias  point  is 
shifted  to  the  right  on  figure  27  until  the  class 
A operating  point  is  reached  at  maximum  ex- 
citation. 

The  bias-shift  voltage  may  be  obtained 
directly  from  the  exciting  signal  by  recti- 
fication, as  shown  in  figure  28.  A simple  low 
pass  filter  system  is  used  that  will  pass  only 
the  syllabic  components  of  speech.  Enough 
negative  bias  is  applied  to  the  bias-shift  modu- 
lator to  cut  the  resting  plate  current  to  the 
desired  value,  and  the  output  of  the  bias  con- 
trol rectifier  is  polarized  so  as  to  ‘*buck** 
the  fixed  bias  voltage.  No  spurious  modu- 
lation frequencies  are  generated,  since  the 
modulator  operates  class  A throughout  the 
audio  cycle. 

This  form  of  grid  pulsing  permits  the  modu- 
lator stage  to  work  with  an  pverall  efficiency 
of  greater  than  50%,  comparing  favorably  with 
the  class  B modulator.  The  expensive  class  B 
driver  and  output  transformers  are  not  required, 
since  resistance  coupling  may  be  used  in  the 
input  circuit  of  the  bias-shift  modulator,  and 


CLASS  C 

AMPLIFIER 


Figure  28 

BLOCK  DIAGRAM  OF 
BIAS-SHIFT  MODULATOR 


a heavy-duty  filter  choke  will  serve  as  an  im- 
pedance coupler  for  the  modulated  stage. 

Series  and  Porallel  The  bias-shift  system 
Control  Circuits  make  take  one  of  several 

forms.  A “series'*  control 
circuit  is  shown  in  figure  29.  Resting  bias  is 
applied  to  the  bias-shift  modulator  tube  through 
the  voltage  divider  R2/R4.  The  bias  control 
tube  is  placed  across  resistor  R2.  Quiescent 
bias  for  the  modulator  is  set  by  adjusting  R2. 
As  the  internal  resistance  of  the  bias  control 
tube  is  varied  at  a syllabic  rate  the  voltage 
drop  across  R2  will  vary  in  unison.  The  modu- 
lator bias,  therefore  varies  at  the  same  rate. 
Excitation  for  the  bias  control  tube  is  obtained 
from  the  audio  signal  through  potentiometer 
R1  which  regulates  the  amplitude  of  the  con- 
trol signal.  The  audio  signal  is  rectified  by 
the  bias  control  rectifier,  and  filtered  by  net- 
work R3-C1  in  the  grid  circuit  of  the  bias  con- 
trol tube. 

The  “parallel**  control  system  is  illustrated 
in  figure  30.  Resting  bias  for  the  modulator  is 
obtained  from  the  voltage  divider  R2/R4. 
Potentiometer  R2  adjusts  the  resting  bias 
level,  determining  the  static  plate  current  of 
the  modulator.  Resistor  R3  serves  as  a bias 
resistor  for  the  control  tube,  reducing  its  plate 
current  to  a low  level.  When  an  audio  signal 
is  applied  via  Rl  to  the  grid  of  the  control 
tube  the  internal  resistance  is  lowered,  de- 
creasing the  shunt  resistance  across  R2.  The 
negative  modulator  bias  is  therefore  reduced. 
The  bias  axis  of  the  modulator  is  shifted  from 
the  cut-off  region  to  a point  on  the  linear 
portion  of  the  operating  curve.  The  amount  of 
bias-shift  is  controlled  by  the  setting  of 
potentiometer  Rl.  Capacitor  Cl  in  conjunction 
with  bias  resistor  R3  form  a syllabic  filter  for 


HANDBOOK 


He i sing  Modulator  309 


B-H  TO  MODULATED 

SPEECH  BIAS-SHIFT  R-F  AMPLIFIER 


NEGATIVE 
MODULATOR  BIAS 


Figure  29 

“SERIES"  CONTROL  CIRCUIT 
FOR  BIAS-SHIFT  MODULATOR 

The  internal  resistance  of  the 
bias  control  tube  is  varied  at  a 
syllabic  rate  to  change  the 
operating  bias  of  the  modulator 
tube. 


the  control  bias  that  is  applied  to  the  modu* 
lator  stage. 

A large  value  of  plate  dissipation  is  re- 
quired for  the  bias-shift  modulator  tube.  For 
plate  voltages  below  1500,  the  211'(VT-4C) 
may  be  used,  while  the  304-TL  is  suitable  for 
voltages  up  to  3000.  As  with  normal  class  A 
amplifiers,  low  mu  tubes  function  best  in  this 
circuit. 


A High-Level  Bias-Shift  Shown  in  figure  31  is 
Modulator  a bias-shift  modulator 

employing  two  304-TL 
tubes  in  parallel,  operating  from  a 2500  volt 
plate  supply.  This  modulator  is  capable  of 
modulating  1000  watts  input  to  an  r-f  amplifier. 
Modulator  resting  current  is  adjusted  to  about 
40  milliamperes,  rising  to  about  350  milliam- 
peres  on  voice  peaks.  Plate  power  requirements 
are  approximately  the  same  as  a conventional 
class  B modulator.  The  class  C r-f  amplifier 
is  adjusted  to  draw  400  milliamperes  at  2500 
volts. 

Two  stages  of  resistance  coupled  speech 
amplification  are  employed,  permitting  the  use 
of  a high  gain  crystal  microphone.  The  6J5 
stage  is  transformer  coupled  to  the  parallel 
connected  304-TL  tubes.  Audio  voltage  to 
drive  the  control  tubes  is  taken  from  the 
secondary  of  the  6J5  coupling  transformer.  A 
dual  6SN7  triode  is  employed  as  the  control 


B+  TO  MODULATED 

SPEECH  BIAS-SHIFT  R-F  AMPLIFIER 


Figure  30 

“PARALLEL"  CONTROL 
CIRCUIT  FOR  BIAS-SHIFT 
MODULATOR 

The  resistance  to  ground  of  point 
A in  the  bias  network  is  varied 
at  a syllabic  rate  by  the  bias 
control  tube. 


rectifier  and  the  control  tube.  A negative  bias 
supply  of  approximately  —430  volts  is  required 
to  reduce  the  plate  current  of  the  304— TL  tubes 
to  the  proper  value.  The  actual  bias  at  the 
grid  of  the  304-TL's  is  -280  volts. 

A separate  filament  supply  is  required  for 
the  6SN7  control  tube  in  order  to  remain  within 
themanufacturer’s  recommended  heater-cathode 
breakdown  voltage.  The  filament  of  this  tube 
is  therefore  connected  to  the  bias  line,  rather 
than  being  grounded  as  are  the  filaments  of 
the  other  6.3  volt  tubes. 

The  resting  bias  potentiometer  is  adjusted 
for  the  proper  quiescent  bias  of  the  304-TL 
tubes,  determined  by  the  static  plate  current 
level,  and  the  audio  bias  control  is  then  ad- 
vanced until  the  operating  bias  of  the  modu- 
lators drops  to  approximately  —210  volts  under 
conditions  of  100%  modulation.  Since  the  modu- 
lator tube  and  the  r-f  amplifier  tube  operate  at 
the  same  plate  voltage,  it  will  not  be  possible 
to  actually  reach  100%  modulation,  but  an 
average  peak  level  of  over  90%  may  easily  be 
obtained. 

A high  level  low-pass  audio  filter  is  placed 
after  the  modulator  stage  to  reduce  the  higher 
harmonic  components  generated  by  the  single- 
ended  amplifiers. 


310 


304-TL  B4- MODULATED 


Figure  31 

500-WATT  BIAS  SHIFT  MODULATOR 

Two  304-TL  triode  tubes  in  bias  controlled  class  A modulator  will 
effectively  modulate  one  kilowatt  input  to  the  class  C amplifier  stage. 
Modulator  tubes  are  operated  at  2500  volts,  and  approximately  —280 
volts  bias.  Resting  bias  is  adjusted  for  static  plate  current  of  40 
milliamperes. 


CHAPTER  SIXTEEN 


Frequency  Modulation  and 
Radioteletype  Transmission 


Exciter  systems  for  FM  aod  single  sideband 
transmission  are  basically  similar  in  that  modi- 
fication of  the  signal  in  accordance  with  the 
intelligence  to  be  transmitted  is  normally  ac- 
complished at  a relatively  low  level.  Then  the 
intelligence-bearing  signal  is  amplified  to  the 
desired  power  level  for  ultimate  transmission. 
True,  amplifiers  for  the  two  types  of  signals 
are  basically  different;  linear  amplifiers  of  the 
Class,  A or  Class  B type  being  used  for  ssb 
signals,  while  Class  C or  non-linear  Class  B 
amplifiers  may  be  used  for  FM  amplification. 
But  the  principle  of  low-level  generation  and 
subsequent  amplification  is  standard  for  both 
types  of  transmission. 

16-1  Frequency  Modulation 

The  use  of  frequency  modulation  and  the 
allied  system  of  phase  modulation  has  become 
of  increasing  importance  in  recent  years.  For 
amateur  communication  frequency  and  phase 
modulation  offer  important  advantages  in  the 
reduction  of  broadcast  and  TV  interference 
and  in  the  elimination  of  the  costly  high-level 
modulation  equipment  most  commonly  employed 
with  amplitude  modulation.  For  broadcast  work 
FM  offers  an  improvement  in  signal-to-noise 
ratio  for  the  high  field  intensities  available 
in  the  local-coverage  area  of  FM  and  TV  broad- 
cast stations. 

In  this  chapter  various  points  of  difference 
between  FM  and  amplitude  modulation  trans- 
mission and  reception  will  be  discussed  and 


the  advantages  of  FM  for  certain  types  of  com- 
munication pointed  out.  Since  the  distinguish- 
ing features  of  the  two  types  of  transmission 
lie  entirely  in  the  modulating  circuits  at  the 
transmitter  and  in  the  detector  and  limiter  cir- 
cuits in  the  receiver,  these  parts  of  the  com- 
munication system  will  receive  the  major  por- 
tion of  attention. 

Modulation  Modulation  is  the  process  of  al- 
tering a radio  wave  in  accordance 
with  the  intelligence  to  be  transmitted.  The 
nature  of  the  intelligence  is  of  little  impor- 
tance as  far  as  the  process  of  modulation  is 
concerned;  it  is  the  method  by  which  this  in- 
telligence is  made  to  give  a distinguishing 
characteristic  to  the  radio  wave  which  will 
enable  the  receiver  to  convert  it  back  into  in- 
telligence that  determines  the  type  of  modu- 
lation being  used. 

Figure  1 is  a drawing  of  an  r-f  carrier  am- 
plitude modulated  by  a sine-wave  audio  volt- 
age. After  modulation  the  resultant  modulated 
r-f  wave  is  seen  still  to  vary  about  the  zero 
axis  at  a constant  rate,  but  the  strength  of  the 
individual  r-f  cycles  is  proportional  to  the  am- 
plitude of  the  modulation  voltage. 

In  figure  2,  the  carrier  of  figure  1 is  shown 
frequency  modulated  by  the  same  modulating 
voltage.  Here  it  may  be  seen  that  modulation 
voltage  of  one  polarity  causes  the  carrier  fre- 
quency to  decrease,  as  shown  by  the  fact  that 
the  individual  r-f  cycles  of  the  carrier  are 
spaced  farther  ^art.  A modulating  voltage  of 
the  opposite  polarity  causes  the  frequency  to 


312 


Frequency  Modulation  313 


FIGURE  1 FIGURE  Z 


AM  AND  FM  WAVES 

Figure  1 shows  a sketch  of  the  scope  pattern  of 
an  awplitude  moduiated  wave  at  the  bottam.  The 
center  sketch  shows  the  modulating  wave  and  the 
upper  sketch  shows  the  carrier  wave. 

Figure  2 shows  at  the  bottom  a sketch  of  a fre- 
quency modulated  wave.  In  this  case  the  center 
sketch  also  shows  the  modulating  wave  and  the 
upper  sketch  shows  the  carrier  wave.  Note  that 
the  carrier  wave  and  the  modulating  wave  are  the 
same  in  either  case,  but  that  the  waveform  of  the 
modulated  wave  is  quite  different  in  the  two 
cases. 


increase,  and  this  is  shown  by  the  r-f  cycles 
being  squeezed  together  to  allow  more  of  them 
to  be  completed  in  a given  time  interval. 

Figures  1 and  2 reveal  two  very  important 
characteristics  about  amplitude-  and  frequen- 
cy-modulated waves.  First,  it  is  seen  that 
while  the  amplitude  (power)  of  the  signal  is 
varied  in  AM  rransmission,  no  such  variation 
takes  place  in  FM.  In  many  cases  this  advan- 
tage of  FM  is  probably  of  equal  or  greater  im- 
portance than  the  widely  publicized  noise  re- 
duction capabilities  of  the  system.  When  100 
per  cent  amplitude  modulation  is  obtained,  the 
average  power  output  of  the  transmitter  must 
be  increased  by  50  per  cent.  This  additional 
output  must  be  supplied  either  by  the  modu- 
lator itself,  in  the  high-level  system,  or  by 
operating  one  or  more  of  the  transmitter  stages 
at  such  a low  output  level  that  they  are  capa- 
ble of  producing  the  additional  output  without 
distortion,  in  the  low-level  system.  On  the 
other  hand,  a frequency-modulated  transmitter 
requires  an  insignificant  amount  of  power  from 
the  modulator  and  needs  no  provision  for  in- 
creased power  output  on  modulation  peaks. 
All  of  the  stages  between  the  oscillator  and 
the  antenna  maybe  operated  as  high-efficiency 
Class  B or  Class  C amplifiers  or  frequency 
multipliers. 


_UNMpDyL^TED_C.^mER  A^PLITl^E 

CARRIER 

UJ 

a 

K SIDE  FREQUENCY  SIDE  FREQUENCY 

CL 

2 

< 

FREQUENCY 

Figure  3 

AM  SIDE  FREQUENCIES 
For  each  AM  modulating  frequency,  a pair  of  side 
frequencies  is  produced.  The  side  frequencies  are 
spaced  away  from  the  carrier  by  an  amount  equal 
to  the  modulation  frequency,  and  their  arrplitude 
is  directly  proportional  to  the  amplitude  of  the 
modulc^ion.  The  amplitude  of  the  carrier  does  not 
change  under  modulation. 


Carrier-Wave  The  second  characteristic  of  FM 
Distortion  and  AM  waves  revealed  by  fig- 
ures 1 and  2 is  that  both  types 
of  modulation  result  in  distortion  of  the  r-f 
carrier.  That  is,  after  modulation,  the  r-f  cy- 
cles are  no  longer  sine  waves,  as  they  would 
be  if  no  frequencies  other  than  the  fundamen- 
tal carrier  frequency  were  present.  It  may  be 
shown  in  the  amplitude  modulation  case  illus- 
trated, that  there  are  only  two  additional  fre- 
quencies present,  and  these  ate  the  familiar 
side  frequencies,  one  located  on  each  side  of 
the  carrier,  and  each  spaced  from  the  carrier 
by  a frequency  interval  equal  to  the  modula- 
tion frequency.  In  regard  to  frequency  and  am- 
plitude, the  situation  is  as  shown  in  figure  3. 
The  strength  of  the  carrier  itself  does  not  vary 
during  modulation,  but  the  strength  of  the  side 
frequencies  depends  upon  the  percentage  of 
modulation.  At  100  per  cent  modulation  the 
power  in  the  side  frequencies  is  equal  to  half 
that  of  the  carrier. 

Under  frequency  modulation,  the  carrier  wave 
again  becomes  distorted,  as  shown  in  figure 
2.  But,  in  this  case,  many  more  than  two  addi- 
tional frequencies  are  formed.  The  first  two  of 
these  frequencies  are  spaced  from  the  carrier 
by  the  modulation  frequency,  and  the  additional 
side  frequencies  are  located  out  on  each  side 
of  the  carrier  and  are  also  spaced  from  each 
other  by  an  amounr  equal  to  the  modulation  fre- 
quency. Theoretically,  there  are  an  Infinite 
number  of  side  frequencies  formed,  but,  for- 
tunately, the  strength  of  those  beyond  the  fre- 
quency swing  of  the  transmitter  under  modula- 
tion is  relatively  low. 

One  set  of  side  frequencies  that  might  be 
formed  by  frequency  modulation  is  shown  in 
figure  4.  Unlike  amplitude  modulation,  the 


314  FM  Transmission 


THE  RADIO 


Figure  4 

FM  SIDE  FREQUENCIES 
With  FM  each  modulation  frequency  component 
causes  o large  number  of  side  frequencies  to  be 
produced.  The  side  frequencies  are  separated 
from  each  other  and  the  carrier  by  an  amount  equal 
to  the  modulation  frequency,  but  their  anpiitude 
varies  greatly  as  the  amount  of  modulation  is 
changed.  The  carrier  strength  also  varies  greatly 
with  frequency  modulation.  The  side  frequencies 
shown  represent  a case  where  the  deviation  each 
side  of  the  '^carrier**  frequency  is  equal  to  five 
times  the  modulating  frequency.  Other  amounts  of 
deviation  with  the  same  modulation  frequency 
would  cause  the  relative  strerrgths  of  the  various 
sidebands  to  chartge  widely. 


strength  of  the  component  at  the  carrier  fre- 
quency varies  widely  in  FM  and  it  may  even 
disappear  entirely  under  certain  conditions. 
The  variation  of  strength  of  the  carrier  com- 
ponent is  useful  in  measuring  the  amount  of 
frequency  modulation,  and  will  be  discussed 
in  detail  later  in  this  chapter. 

One  of  the  great  advantages  of  FM  over  AM 
is  the  reduction  in  noise  at  the  receiver  which 
the  system  allows.  If  the  receiver  is  made  re- 
sponsive only  to  changes  in  frequency,  a con- 
siderable increase  in  signal-to-noise  ratio  is 
made  possible  through  the  use  of  FM,  when 
the  signal  is  of  greater  strength  than  the  noise. 
The  noise  reducing  capabilities  of  FM  arise 
from  the  inability  of  noise  to  cause  appreciable 
frequency  modulation  of  the  noise-plus-signal 
voltage  which  is  applied  to  the  detector  in  the 
receiver. 

FM  Terms  Unlike  amplitude  modulation,  the 
term  percentage  modulation  means 
little  in  FM  practice,  unless  the  receiver  char- 
acteristics are  specified.  There  are,  however, 
three  terms,  deviation,  modulation  index,  and 
deviation  ratio,  which  convey  considerable 
information  concerning  the  character  of  the 
FM  wave. 

Deviation  is  the  amount  of  frequency  shift 
each  side  of  the  unmodulated  carrier  frequency 
which  occurs  when  the  transmitter  is  modu- 
lated. Deviation  is  ordinarily  measured  in  kilo- 
cycles, and  in  a properly  operating  FM  trans- 


mitter it  will  be  directly  proportional  to  the 
amplitude  of  the  modulating  signal.  When  a 
symmetrical  modulating  signal  is  applied  to 
the  transmitter,  equal  deviation  each  side  of 
the  resting  frequency  is  obtained  during  each 
cycle  of  the  modulating  signal,  and  the  total 
frequency  range  covered  by  the  FM  transmitter 
is  sometimes  known  as  the  swing.  If,  for  in- 
stance, a transmitter  operating  on  1000  kc. 
has  its  frequency  shifted  from  1000  kc.  to  1010 
kc.,  back  to  1000  kc.,  then  to  990  kc.,  and 
again  back  to  1000  kc.  during  one  cycle  of  the 
modulating  wave,  the  deviation  would  be  10 
kc.  and  the  swing  20  kc. 

The  modulation  index  of  an  FM  signal  is 
the  ratio  of  the  deviation  to  the  audio  modu- 
lating frequency,  when  both  are  expressed  in 
the  same  units.  Thus,  in  the  example  above 
if  the  signal  is  varied  from  1000  kc.  to  1010 
kc.  to  990  kc.,  and  back  to  1000  kc.  at  a rate 
(frequency)  of  2000  times,  a second,  the  mod- 
ulation index  would  be  5,  since  the  deviation 
(10  kc.)  is  5 times  the  modulating  frequency 
(2000  cycles,  or  2 kc.). 

The  relative  strengths  of  the  FM  carrier  and 
the  various  side  frequencies  depend  directly 
upon  the  modulation  index,  these  relative 
strengths  varying  widely  as  the  modulation 
index  is  varied.  In  the  preceding  example,  for 
instance,  side  frequencies  occur  on  the  high 
side  of  1000  kc.  at  1002,  1004,  1006,  1008, 
1010,  1012,  etc.,  and  on  the  low  frequency 
side  at  998,  996,  994,  992,  990,  988,  etc.  In 
proportion  to  the  unmodulated  carrier  strength 
(100  per  cent),  these  side  frequencies  have 
the  following  strengths,  as  indicated  by  a 
modulation  index  of  5:  1002  and  998—33  per 
cent,  1004  and  996—5  per  cent,  1006  and  994— 
36  per  cent,  1008  and  992—39  pet  cent,  1010 
and  990—26  per  cent,  1012  and  988—13  per 
cent.  The  carrier  strength  (1000  kc.)  will  be 
18  per  cent  of  its  unmodulated  value.  Chang- 
ing the  amplitude  of  the  modulating  signal  will 
change  the  deviation,  and  thus  the  modulation 
index  will  be  changed,  with  the  result  that  the 
side  frequencies,  while  still  located  in  the 
same  places,  will  have  different  strength  values 
from  those  given  above. 

The  deviation  ratio  is  similar  to  the  modu- 
lation index  in  that  it  involves  the  ratio  be- 
tween a modulating  frequency  and  deviation. 
In  this  case,  however,  the  deviation  in  ques- 
tion is  the  peak  frequency  shift  obtained  under 
full  modulation,  and  the  audio  frequency  to  be 
considered  is  the  maximum  audio  frequency  to 
be  transmitted.  When  the  maximum  audio  fre- 
quency to  be  transmitted  is  5000  cycles,  for 
example,  a deviation  ratio  of  3 would  call  for 
a peak  deviation  of  3 x 5000,  or  15  kc.  at  full 
modulation.  The  noise-suppression  capabili- 
ties of  FM  are  directly  related  to  the  devia- 
tion ratio.  As  the  deviation  ratio  is  increased. 


HANDBOOK 


Narrow  Band  FM  315 


the  noise  suppression  becomes  better  if  the 
signal  is  somewhat  stronger  than  the  noise. 
Where  the  noise  approaches  the  signal  in 
strength,  however,  low  deviation  ratios  allow 
communication  to  be  maintained  in  many  cases 
where  high-deviation-ratio  FM  and  conven- 
tional AM  ate  incapable  of  giving  service. 
This  assumes  that  a narrow-band  FM  receiver 
is  in  use.  For  each  value  of  t-f  signal-to-noise 
ratio  at  the  receiver,  there  is  a maximum  de- 
viation ratio  which  may  be  used,  beyond  which 
the  output  audio  signal-to-noise  ratio  de- 
creases. Up  to  this  critical  deviation  ratio, 
however,  the  noise  suppression  becomes  pro- 
gressively better  as  the  deviation  ratio  is  in- 
creased. 

For  high-fidelity  FM  broadcasting  purposes, 
a deviation  ratio  of  5 is  ordinarily  used,  the 
maximum  audio  frequency  being  15,000  cycles, 
and  the  peak  deviation  at  full  modulation  be- 
ing 75  kc.  Since  a swing  of  150  kc.  is  coveted 
by  the  transmitter,  it  is  obvious  that  wide- 
band FM  transmission  must  necessarily  be 
confined  to  the  v-h-f  range  or  higher,  where 
room  for  the  signals  is  available. 

In  the  case  of  television  sound,  the  devia- 
tion ratio  is  1.67;  the  maximum  modulation 
frequency  is  15,000  cycles,  and  the  trans- 
mitter deviation  for  full  modulation  is  25  kc. 
The  sound  carrier  frequency  in  a standard  TV 
signal  is  located  exactly  4.5  Me.  higher  than 
the  picture  carrier  frequency.  In  the  infer- 
carrier  TV  sound  system,  which  recently  has 
become  quite  widely  used,  this  constant  differ- 
ence between  the  picture  carrier  and  the  sound 
carrier  is  employed  within  the  receiver  to  ob- 
tain an  FM  sub-carrier  at  4.5  Me.  This  4.5 
Me.  sub-carrier  then  is  demodulated  by  the  FM 
detector  to  obtain  the  sound  signal  which 
accompanies  the  picture. 

Narrow- Band  Narrow-band  FM  trans- 

FM  Transmission  mission  has  become  stand- 
ardized for  use  by  the  mo- 
bile services  such  as  police,  fire,  and  taxi- 
cab communication,  and  also  on  the  basis  of 
a temporary  authorization  for  amateur  work  in 
portions  of  each  of  the  amateur  radiotelephone 
bands.  A maximum  deviation  of  15  kc.  has 
been  standardized  for  the  mobile  and  commer- 
cial communication  services,  while  a maxi- 
mum deviation  of  3 kc.  is  authorized  for  ama- 
teur NBFM  communication. 

Bandwidth  Re-  As  the  above  discussion  has 
quired  by  FM  indicated,  many  side  frequen- 
cies are  set  up  when  a radio- 
frequency carrier  is  frequency  modulated;  theo- 
retically, in  fact,  an  infinite  number  of  side 
frequencies  is  formed.  Fortunately,  however, 
the  amplitudes  of  those  side  frequencies  fall- 
ing outside  the  frequency  range  over  which 


the  transmitter  is  swung  are  so  small  that 
most  of  them  may  be  ignored.  In  FM  trans- 
mission, when  a complex  modulating  wave 
(speech  or  music)  is  used,  still  additional 
side  frequencies  resulting  from  a beating  to- 
gether of  the  various  frequency  components 
in  the  modulating  wave  ate  formed.  This  is  a 
situation  that  does  not  occur  in  amplitude 
modulation  and  it  might  be  thought  that  the 
large  number  of  side  frequencies  thus  formed 
might  make  the  frequency  spectrum  produced 
by  an  FM  transmitter  prohibitively  wide.  Analy- 
sis shows,  however,  that  the  additional  side 
frequencies  are  of  very  small  amplitude,  and, 
instead  of  increasing  the  bandwidth,  modula- 
tion by  a complex  wave  actually  reduces  the 
effective  bandwidth  of  the  FM  wave.  This  is 
especially  true  when  speech  modulation  is 
used,  since  most  of  the  power  in  voiced  sounds 
is  concentrated  at  low  frequencies  in  the  vicin- 
ity of  400  cycles. 

The  bandwidth  required  in  an  FM  receiver 
is  a function  of  a number  of  factors,  both  theo- 
retical and  practical.  Basically,  the  bandwidth 
required  is  a function  of  the  deviation  ratio 
and  the  maximum  frequency  of  modulation, 
although  the  practical  consideration  of  drift 
and  ease  of  receiver  tuning  also  must  be  con- 
sidered. Shown  in  figure  5 are  the  frequency 
spectra  (carrier  and  sideband  frequencies) 
associated  with  the  standard  FM  broadcast 
signal,  the  TV  sound  signal,  and  an  amateur- 
band  narrow-band  FM  signal  with  full  modula- 
tion using  the  highest  permissible  modulating 
frequency  in  each  case.  It  will  be  seen  that 
for  low  deviation  ratios  the  receiver  band- 
width should  be  at  least  four  times  the  maxi- 
mum frequency  deviation,  but  for  a deviation 
ratio  of  5 the  receiver  bandwidth  need  be  only 
about  2.5  times  the  maximum  frequency  de- 
viation. 

16-2  Direct  FM  Circuits 

Frequency  modulation  may  be  obtained  either 
by  the  direct  method,  in  which  the  frequency 
of  an  oscillator  is  changed  directly  by  the 
modulating  signal,  or  by  the  indirect  method 
which  makes  use  of  phase  modulation.  Phase- 
modulation  circuits  will  be  discussed  in  sec- 
tion 16-3. 

A successful  frequency  modulated  trans- 
mitter must  meet  two  requirements;  (1)  The 
frequency  deviation  must  be  symmetrical  about 
a fixed  frequency,  for  symmetrical  modulation 
voltage.  (2)  The  deviation  must  be  directly 
proportional  to  the  amplitude  of  the  modula- 
tion, and  independent  of  the  modulation  fre- 
quency. There  are  several  methods  of  direct 
frequency  modulation  which  will  fulfill  these 


316  FM  Transmission 


THE  RADIO 


FM  BROADCAST  DEVIATION  - 75  RC. 

MOO.  FREQ.- 15  KC. 
MOD.  INDEX  - 5 


# ^ 1?  1-  1- 

Im  |o>  I'O 

(N)  U |(M  In  |n 

<£ 

i \Af.  % 

- 1#  1#  ^ 

^ 1*^  1-  I'-  - 

^ |<D  loi  ltd  |rfi 

-105  -90  -75  -60  -45  -30  -15 

+ 15  +30  +45  +60  +75  +90  +105 

(b)  tv  sound 

OEVIATlON-25  KC. 

. MOO.  FREQ.- 15  KC. 

' MOD.  INDEX  -1.67 

1 

f f 

S 15 

II-  |n 

5 7 Vo 

ICARRIER 

# 

+ 

1 i * 

¥■  I*-  ■f' 

r 1^  1^ 

-45  KC.  -30KC.  - 

15  KC. 

+ 15KC.  -I-30KC.  -F45KC. 

@ AMATEUR  NBFM 

1 

DEVIATION  - 3 KC. 

MOO.  FREQ-  3 KC. 

MOO.  INDEX  - 1 

CARRIER 

(b 

If  J 

-6KC.  -3KC. 

1 

1 

4-  3 KC.  +6  KC. 

1 

CENTER 

FREQUENCY 

Figure  5 

EFFECT  OF  FM  MODULATION  INDEX 
Showing  the  side- frequency  amplitude  and  distri- 
bution for  the  three  rnost  common  modulation  indi- 
ces used  in  FM  work.  The  maximum  modulating 
frequency  and  maximum  deviation  are  shown  in 
each  case. 

requirements.  Some  of  these  methods  will  be 
described  in  the  following  paragraphs. 

Reactance-Tube  One  of  the  most  practical 
Modulators  ways  of  obtaining  direct  fre- 

quency modulation  is  through 
the  use  of  a reactance-tube  modulator.  In  this 
arrangement  the  modulator  plate-cathode  cir- 
cuit is  connected  across  the  oscillator  tank 
circuit,  and  made  to  appear  as  either  a capaci- 
tive or  inductive  reactance  by  exciting  the 
modulator  grid  with  a voltage  which  either 
leads  or  lags  the  oscillator  tank  voltage  by 
90  degrees.  The  leading  or  lagging  grid  volt- 
age causes  a corresponing  leading  or  lagging 
plate  current,  and  the  plate-cathode  circuit 
appears  as  a capacitive  or  inductive  reactance 
across  the  oscillator  tank  circuit.  When  the 
trans conductance  of  the  modulator  tube  is 
varied,  by  varying  one  of  the  element  voltages, 
the  magnitude  of  the  reactance  across  the  os- 


in  OSCILLATOR  IN 


This  circuit  is  convenient  for  direct  frequen- 
cy modulation  of  an  oscillator  in  the  1.75-Mc. 
ronge.  Capacitor  C3  may  be  only  the  input 
capacitance  of  the  tube,  or  o small  trimmer 
capacitor  may  be  included  to  permit  a varia- 
tion in  the  sensitivity  of  the  reactance  tube. 


cillator  tank  is  varied.  By  applying  audio  mod- 
ulating voltage  to  one  of  the  elements,  the 
transconductance,  and  hence  the  frequency, 
may  be  varied  at  an  audio  rate.  When  properly 
designed  and  operated,  the  reactance-tube 
modulator  gives  linear  frequency  modulation, 
and  is  capable  of  producing  large  amounts  of 
deviation. 

There  are  numerous  possible  configurations 
of  the  reactance-tube  modulator  circuit.  The 
difference  in  the  various  arrangements  lies 
principally  in  the  type  of  phase-shifting  cir- 
cuit used  to  give  a grid  voltage  which  is  in 
phase  quadrature  with  the  t-f  voltage  at  the 
modulator  plate. 

Figure  6 is  a diagram  of  one  of  the  most 
popular  forms  of  reactance-tube  modulators. 
The  modulator  tube,  which  is  usually  a pen- 
tode such  as  a 6BA6,  6AU6,  or  6CL6,  has  its 
plate  coupled  through  a blocking  capacitor, 
C,,  to  the  "hot”  side  of  the  oscillator  grid 
circuit.  Another  blocking  capacitor,  Cj,  feeds 
r.f.  to  the  phase  shifting  network  R-Cj  in  the 
modulator  grid  circuit.  If  the  resistance  of  R 
is  made  large  in  comparison  with  the  react- 
ance of  C3  at  the  oscillator  frequency,  the  cur- 
rent through  the  R-C3  combination  will  be 
nearly  in  phase  with  the  voltage  across  the 
tank  circuit,  and  the  voltage  across  C3  will 
lag  the  oscillator  tank  voltage  by  almost  90 
degrees.  The  result  of  the  90-degree  lagging 
voltage  on  the  modulator  grid  is  that  its  plate 
current  lags  the  tank  voltage  by  90  degrees, 
and  the  reactance  tube  appears  as  an  induct- 
ance in  shunt  with  the  oscillator  inductance, 
thus  raising  the  oscillator  frequency. 

The  phase-shifting  capacitor  C3  can  consist 
of  the  input  capacitance  of  the  modulator  tube 
and  stray  capacitance  between  grid  and  ground. 


HANDBOOK 


Reactance  Tube  317 


Figure  7 

ALTERNATIVE  REACTANCE-TUBE 
MODULATOR 

This  circuit  is  often  preferable  for  use  in  the 
tower  frequency  range,  although  it  may  be 
used  at  1,75  Me.  and  above  if  desired.  In  the 
schematic  above  the  reactance  tube  is  shown 
connected  across  the  voltage^divider  capaci- 
tors  of  a Clapp  oscillator,  although  the  modu‘ 
lator  circuit  may  be  used  with  any  common 
type  of  oscillator. 


However,  better  control  of  the  operating  con- 
ditions of  the  modulator  may  be  had  through 
the  use  of  a variable  capacitor  as  Cj.  Resist- 
ance R will  usually  have  a value  of  between 
4700  and  100,000  ohms.  Either  resistance  or 
transformer  coupling  may  be  used  to  feed  audio 
voltage  to  the  modulator  grid.  When  a resist- 
ance coupling  is  used,  it  is  necessary  to  shield 
the  grid  circuit  adequately,  since  the  high  im- 
pedance grid  circuit  is  prone  to  pick  up  stray 
r-f  and  low  frequency  a-c  voltage,  and  cause 
undesired  frequency  modulation. 

An  alternative  reactance  modulator  circuit 
is  shown  in  figure  7.  The  operating  conditions 
ate  generally  the  same,  except  that  the  r-f 
excitation  voltage  to  the  grid  of  the  reactance 
tube  is  obtained  effectively  through  reversing 
the  R and  Cj  of  figure  6.  In  this  circuit  a small 
capacitance  is  used  to  couple  r.f.  into  the  grid 
of  the  reactance  tube,  with  a relatively  small 
value  of  resistance  from  grid  ro  ground.  This 
circuit  has  the  advantage  that  the  grid  of  the 
tube  is  at  relatively  low  impedance  with  re- 
spect to  r.f.  However,  the  circuit  normally  is 
not  suitable  for  operation  above  a few  mega- 
cycles due  to  the  shunting  capacitance  within 
the  tube  from  grid  to  ground. 

Either  of  the  reactance-tube  circuits  may 
be  used  with  any  of  the  common  types  of  os- 
cillators. The  reactance  modulator  of  figure 
6 is  shown  connected  to  the  high-impedance 
point  of  a conventional  hot-cathode  Hartley 
oscillator,  while  that  of  figure  7 is  shown  con- 
nected across  the  low-impedance  capacitors 

of  a series-tuned  Clapp  oscillator. 

There  are  several  possible  variations  of  the 
basic  reactance-tube  modulator  circuits  shown 


in  figures  6 and  7.  The  audio  input  may  be 
applied  to  the  suppressor  grid,  rather  than  the 
control  grid,  if  desired.  Another  modification 
is  to  apply  the  audio  to  a grid  other  than  the 
control  grid  in  a mixer  or  pentagrid  converter 
tube  which  is  used  as  the  modulator.  Gener- 
ally, it  will  be  found  that  the  transconductance 
variation  per  volt  of  control-element  voltage 
variation  will  be  greatest  when  the  control 
(audio)  voltage  is  applied  to  the  control  grid. 
In  cases  where  it  is  desirable  to  separate  com- 
pletely the  audio  and  r-f  circuits,  however, 
applying  audio  voltage  to  one  of  the  other  ele- 
ments will  often  be  found  advantageous  de- 
spite the  somewhat  lower  sensitivity. 

Adjusting  the  One  of  the  simplest  methods 
Phase  Shift  of  adjusting  the  phase  shift  to 
the  correct  amount  is  to  place 
a pair  of  earphones  in  series  with  the  oscilla- 
tor cathode-to- ground  circuit  and  adjust  the 
phase-shift  network  until  minimum  sound  is 
heard  in  the  phones  when  frequency  modula- 
tion is  taking  place.  If  an  electron-coupled  or 
Hartley  oscillator  is  used,  this  method  re- 
quires that  the  cathode  circuit  of  the  oscilla- 
tor be  inductively  or  capacitively  coupled  to 
the  grid  circuit,  rather  than  tapped  on  the  grid 
coil.  The  phones  should  be  adequately  by- 
passed for  r.f.  of  course. 

Stabilization  Due  to  the  presence  of  the  react- 
ance-tube frequency  modulator, 
the  stabilization  of  an  FM  oscillator  in  regard 
to  voltage  changes  is  considerably  more  in- 
volved than  in  the  case  of  a simple  self-con- 
trolled oscillator  for  transmitter  frequency 
control.  If  desired,  the  oscillator  itself  may  be 
made  perfectly  stable  under  voltage  changes, 
but  the  presence  of  the  frequency  modulator 
destroys  the  beneficial  effect  of  any  such 
stabilization.  It  thus  becomes  desirable  to 
apply  the  stabilizing  arrangement  to  the  modu- 
lator as  well  as  the  oscillator.  If  the  oscilla- 
tor itself  is  stable  under  voltage  changes,  it 
is  only  necessary  to  apply  voltage-frequency 
compensation  to  the  modulator. 

Reactance-Tube  Two  simple  reactance-tube 
Modulatars  modulators  that  maybe  applied 

to  an  existing  v.f.o.  are  illus- 
trated in  figures  8 and  9.  The  circuit  of  figure 
8 is  extremely  simple,  yet  effective.  Only  two 
tubes  ate  used  exclusive  of  the  voltage  regu- 
lator tubes  which  perhaps  may  be  already  in- 
corporated in  the  v.f.o.  A 6AU6  serves  as  a 
high-gain  voltage  amplifier  stage,  and  a 6CL6 
is  used  as  the  reactance  modulator  since  its 

high  value  of  transconductance  will  permit  a 

large  value  of  lagging  current  to  be  drawn 
under  modulation  swing.  The  unit  should  be 


318  FM  Transmission 


THE  RADIO 


6AU6  6CL6 


Figure  8 

SIMPLE  FM  REACTANCE-TUBE  MODULATOR 


mounted  in  close  proximity  to  the  v.f.o.  so  that 
the  lead  from  the  6CL6  to  the  grid  circuit  of 
the  oscillator  can  be  as  short  as  possible.  A 
practical  solution  is  to  mount  the  reactance 
modulator  in  a small  box  on  the  side  of  the 
v-f-o  cabinet. 

By  incorporating  speech  clipping  in  the  re- 
actance modulator  unit,  a much  more  effective 
use  is  made  of  a given  amount  of  deviation. 
When  the  FM  signal  is  received  on  an  AM  re- 
ceiver by  means  of  slope  detection,  the  use 
of  speech  clipping  will  be  noticed  by  the  great- 
ly increased  modulation  level  of  the  FM  sig- 
nal, and  the  attenuation  of  the  center  frequency 
null  of  no  modulation.  In  many  cases,  it  is 
difficult  to  tell  a speech-clipped  FM  signal 
from  the  usual  AM  signal. 

A mote  complex  FM  reactance  modulator  in- 
corporating a speech  clipper  is  shown  in  fig- 
ure 9.  A 12AX7  double  triode  speech  amplifier 
provides  enough  gain  for  proper  clipper  action 
when  a high  level  crystal  microphone  is  used. 
A double  diode  6AL5  speech  clipper  is  used, 
the  clipping  level  being  set  by  the  potentiome- 
ter controlling  the  plate  voltage  applied  to  the 
diode.  A 6CL6  serves  as  the  reactance  modu- 
lator. 


The  reactance  modulator  may  best  be  ad- 
justed by  listening  to  the  signal  of  the  v-f-o 
exciter  at  the  operating  frequency  and  adjust- 
ing the  gain  and  clipping  controls  for  the  best 
modulation  level  consistent  with  minimum  side- 
band splatter.  Minimum  clipping  occurs  when 
the  Adj.  Clip,  potentiometer  is  set  for  maximum 
voltage  on  the  plates  of  the  6AL5  clipper  tube. 
As  with  the  case  of  all  reactance  modulators, 
a voltage  regulated  plate  supply  is  required. 


Linearity  Tait  It  is  almost  a necessity  to  run 
a static  test  on  the  reactance- 
tube  frequency  modulator  to  determine  its  line- 
arity and  effectiveness,  since  small  changes 
in  the  values  of  components,  and  in  stray  ca- 
pacitances will  almost  certainly  alter  the  modu- 
lator characteristics.  A frequency-vetsus-con- 
trol-voltage  curve  should  be  plotted  to  ascer- 
tain that  equal  increments  in  control  voltage, 
both  in  a positive  and  a negative  direction, 
cause  equal  changes  in  frequency.  If  the  curve 
shows  that  the  modulator  has  an  appreciable 
amount  of  non-linearity,  changes  in  bias,  elec- 
trode voltages,  r-f  excitation,  and  resistance 


HANDBOOK 


Phase  Modulation 


319 


TO  MODULATOR 
CONTROL  ELEMENT 


Figure  10 

REACTANCE-TUBE  LINEARITY  CHECKER 


values  may  be  made  to  obtain  a straight-line 
characteristic. 

Figure  10  shows  a method  of  connecting 
two  4^2- volt  C batteries  and  a potentiometer 
to  plot  the  characteristic  of  the  modulator.  It 
will  be  necessary  to  use  a zero-center  volt- 
meter to  measure  the  grid  voltage,  or  else  re- 
verse the  voltmeter  leads  when  changing  from 
positive  to  negative  grid  voltage.  When  a 
straight-line  characteristic  for  the  modulator 
is  obtained  by  the  static  test  method,  the  ca- 
pacitances of  the  various  by-pass  capacitors 
in  the  circuit  must  be  kept  small  to  retain  this 
charactetistic  when  an  audio  voltage  is  used 
to  vary  the  frequency  in  place  of  the  d-c  volt- 
age with  which  the  characteristic  was  plotted. 

16-3  Phase  Modulation 

By  means  of  phase  modulation  (PM)  it  is 
possible  to  dispense  with  self-controlled  os- 
cillators and  to  obtain  directly  crystal-con- 
trolled  FM.  In  the  fintd  analysis,  PM  is  sim- 
ply frequency  modulation  in  which  the  devia- 
tion is  directly  proportional  to  the  modulation 
frequency.  If  an  audio  signal  of  1000  cycles 
causes  a deviation  of  kc.,  for  example,  a 
2000-cycle  modulating  signal  of  the  same  am- 
plitude will  give  a deviation  of  1 kc.,  and  so 
on.  To  produce  an  FM  signal,  it  is  necessary 
to  make  the  deviation  independent  of  the  modu- 
lation frequency,  and  proportional  only  to  the 
modulating  signal.  With  PM  this  is  done  by 
including  a frequency  correcting  network  in 
the  transmitter.  The  audio  correction  network 
must  have  an  attenuation  that  varies  directly 
with  frequency,  and  this  requirement  is  easily 
met  by  a very  simple  resistance-capacity  net- 
work. 

The  only  disadvantage  of  PM,  as  compared 
to  direct  FM  such  as  is  obtained  through  the 
use  of  a reactance-tube  modulator,  is  the  fact 
that  very  little  frequency  deviation  is  pro- 
duced directly  by  the  phase  modulator.  The 
deviation  produced  by  a phase  modulator  is 
independent  of  the  actual  carrier  frequency  on 


which  the  modulator  operates,  but  is  depend- 
ent only  upon  the  phase  deviation  which  is 
being  produced  and  upon  the  modulation  fre- 
quency. Expressed  as  an  equation: 

F j ■=  Mp  modulating  frequency 

Where  Fj  is  the  frequency  deviation  one  way 
from  the  mean  value  of  the  carrier,  and  Mp  is 
the  phase  deviation  accompanying  modulation 
expressed  in  radians  (a  radian  is  approximately 
57.3°).  Thus,  to  take  an  example,  if  the  phase 
deviation  is  M radian  and  the  modulating  fre- 
quency is  1000  cycles,  the  frequency  deviation 
applied  to  the  carrier  being  passed  through 
the  phase  modulator  will  be  500  cycles. 

It  is  easy  to  see  that  an  enormous  amount 
of  multiplication  of  the  carrier  frequency  is 
required  in  order  to  obtain  from  a phase  modu- 
lator the  frequency  deviation  of  75  kc.  requited 
for  commercial  FM  broadcasting.  However,  for 
amateur  and  commercial  narrow-band  FM  work 
(NBFM)  only  a quite  reasonable  number  of 
multiplier  stages  are  required  to  obtain  a de- 
viation ratio  of  approximately  one.  Actually, 
phase  modulation  of  approximately  one-half 
radian  on  the  output  of  a crystal  oscillator  in 
the  80-metet  band  will  give  adequate  deviation 
for  29-Mc.  NBFM  radiotelephony.  For  example; 
if  the  crystal  frequency  is  3700  kc.,  the  de- 
viation in  phase  produced  is  'A  radian,  and  the 
modulating  frequency  is  500  cycles,  the  devia- 
tion in  the  80-meter  band  will  be  250  cycles. 
But  when  the  crystal  frequency  is  multiplied 
on  up  to  29,600  kc.  the  frequency  deviation 
will  also  be  multiplied  by  8 so  that  the  result- 
ing deviation  on  the  10-meter  band  will  be  2 kc. 
either  side  of  the  carrier  for  a total  swing  in 
carrier  frequency  of  4 kc.  This  amount  of  de- 
viation is  quite  adequate  for  NBFM  work. 

Odd-harmonic  distortion  is  produced  when 
FM  is  obtained  by  the  phase-modulation  meth- 
od, and  the  amount  of  this  distortion  that  can 
be  tolerated  is  the  limiting  factor  in  determin- 
ing the  amount  of  PM  that  can  be  used.  Since 
the  aforementioned  frequency-correcting  net- 
work causes  the  lowest  modulating  frequency 
to  have  the  greatest  amplitude,  maximum  phase 
modulation  takes  place  at  the  lowest  modu- 
lating frequency,  and  the  amount  of  distortion 
that  can  be  tolerated  at  this  frequency  deter- 
mines the  maximum  deviation  that  can  be  ob- 
tained by  the  PM  method.  For  high-fidelity 
broadcasting,  the  deviation  produced  by  PM  is 
limited  to  an  amount  equal  to  about  one-third 
of  the  lowest  modulating  frequency.  But  for 
NBFM  work  the  deviation  may  be  as  high  as 
0.6  of  the  modulating  frequency  before  distor- 
tion becomes  objectionable  on  voice  modula- 
tion. In  Other  terms  jhis  means  that  phase  de- 
viations as  high  as  0.6  radian  may  be  used  for 
amateur  and  commercial  NBFM  transmission. 


320  FM  Transmission 


THE  RADIO 


REACTANCE-TUBE  MODULATION  OF 
CRYSTAL  OSCILLATOR  STAGE 


Phase-Modulation  A simple  reactance  modula- 
Circuits  tor  normally  used  for  FM 

may  also  be  used  for  PM  by 
connecting  it  to  the  plate  citcuit  of  a crystal 
oscillator  stage  as  shown  in  figure  11. 

Another  PM  circuit,  suitable  for  operation 
on  20,  15  and  10  meters  with  the  use  of  80 
meter  crystals  is  shown  in  figure  12.  A double 
triode  12AX7  is  used  as  a combination  Pierce 
crystal  oscillator  and  phase  modulator.  C, 
should  not  be  thought  of  as  a neutralizing  con- 
denser, but  rather  as  an  adjustment  for  the 
phase  of  the  r-f  voltage  acting  between  the 
grid  and  plate  of  the  12AX7  phase  modulator. 
C]  acts  as  a phase  angle  and  magnitude  con- 
trol, and  both  these  condensets  should  be  ad- 
justed for  maximum  phase  modulation  capabili- 
ties of  the  circuit.  Resonance  of  the  circuit  is 
established  by  the  iron  slug  of  coil  Lj-Lj.  A 
6CL6  is  used  as  a doubler  to  7 Me.  and  de- 
livers ^proximately  2 watts  on  this  band.  Ad- 
ditional doublet  stages  may  be  added  after  the 
6CL6  stage  to  reach  the  desired  hand  of  opera- 
tion. 

Still  another  PM  circuit,  which  is  quite  wide- 
ly used  commercially,  is  shown  in  figure  13. 
In  this  circuit  L and  C are  made  resonant  at  a 
frequency  which  is  0.707  times  the  operating 
frequency.  Hence  at  the  operating  frequency 
the  inductive  reactance  is  twice  the  capacitive 
reactance.  A cathode  follower  tube  acts  as  a 
variable  resistance  in  series  with  the  L and 
C which  go  to  make  up  the  tank  citcuit.  The 
operating  point  of  the  cathode  follower  should 
be  chosen  so  that  the  effective  resistance  in 
series  with  the  tank  circuit  (made  up  of  the 
resistance  of  the  cathode-follower  tube  in  par- 
allel with  the  cathode  bias  resistor  of  the  cath- 
ode follower)  is  equal  to  the  capacitive  react- 
ance of  the  tank  capacitor  at  the  operating  fre- 
quency. The  circuit  is  capable  of  about  plus 
or  minus  H radian  deviation  with  tolerable  dis- 
tortion. 


Measurement  When  a single-frequency  mod- 

of  Deviation  ulating  voltage  is  used  with  an 
FM  transmitter,  the  relative 
amplitudes  of  the  various  sidebands  and  the 
carrier  vary  widely  as  the  deviation  is  varied 
by  increasing  or  decreasing  tbe  amount  of  mod- 
ulation. Since  tbe  relationship  between  the 
amplitudes  of  the  various  sidebands  and  car- 
rier to  the  audio  modulating  frequency  and  the 
deviation  is  known,  a simple  method  of  meas- 
uring the  deviation  of  a frequency  modulated 
transmitter  is  possible.  In  making  the  meas- 
urement, the  result  is  given  in  the  form  of  the 
modulation  index  for  a certain  amount  of  audio 
input.  As  previously  described,  the  modulation 
index  is  the  ratio  of  the  peak  frequency  devia- 
tion to  the  frequency  of  the  audio  modulation. 

The  measurement  is  made  by  applying  a 
sine-wave  audio  voltage  of  known  frequency 
to  the  transmitter,  and  increasing  the  modula- 
tion until  the  amplitude  of  the  carrier  compo- 
nent of  the  frequency  modulated  wave  teaches 
zero.  The  modulation  index  for  zero  carrier 
may  then  be  determined  from  the  table  below. 
As  may  be  seen  from  the  table,  the  first  point 
of  zero  carrier  is  obtained  when  the  modulation 
index  has  a value  of  2.405,— in  other  words, 
when  the  deviation  is  2.405  times  the  modula- 
tion frequency.  For  example,  if  a modulation 
frequency  of  1000  cycles  is  used,  and  the 
modulation  is  increased  until  the  first  carrier 
null  is  obtained,  the  deviation  will  then  be 
2.405  times  the  modulation  frequency,  or  2.405 
kc.  If  the  modulating  frequency  happened  to  be 
2000  cycles,  the  deviation  at  the  first  null 
would  be  4.810  kc.  Other  carrier  nulls  will  be 
obtained  when  the  index  is  5.52,  8.654,  and  at 
increasing  values  separated  approximately  by 
n.  The  following  is  a listing  of  the  modulation 
index  at  successive  carrier  nulls  up  to  the 
tenth: 


Zero  carrier 
point  no. 

1 

2 

3 

4 

5 

6 

7 

8 

9 

10 


Modulation 

index 

2.405 

5.520 

8.654 

11.792 

14.931 

18.071 

21.212 

24.353 

27.494 

30.635 


The  only  equipment  required  for  making  the 
measurements  is  a calibrated  audio  oscillator 
of  good  wave  form,  and  a communication  re- 
ceiver equipped  with  a beat  oscillator  and 
crystal  filter.  The  receiver  should  be  used 
with  its  crystal  filter  set  for  minimum  band- 
width to  exclude  sidebands  spaced  from  the 
carrier  hy  the  modulation  frequency.  The  un- 


HANDBOOK 


FM  Reception  321 


I2AX7  6CL6 


-►TO  DOUBLER  STAGES 


L I -35r.  #3e£,\  SPACED  4"apart  on 4' form 

L2~I5T-  #36£.J  POWDERBD  IRON  CORE 
L3-377:  itZOE.  CLOSE-SPACED  ^ DIA. 


NOTE;  ALL  RESISTORS  0.5WATT  UNLESS 
OTHERWISE  NOTED. 

ALL  CAPACITORS  INUJF  UNLESS 
OTHERWISE  NOTED 


Figure  12 

REACTANCE  MODULATOR  FOR  10,  15  AND  20  METER  OPERATION 


modulated  carrier  is  accurately  tuned  in  on  the 
receiver  with  the  beat  oscillator  operating. 
Then  modulation  from  the  audio  oscillator  is 
applied  to  the  transmitter,  and  the  modulation 
is  increased  until  the  first  carrier  null  is  ob- 
tained. This  carrier  null  will  correspond  to  a 
modulation  index  of  2.405,  as  previously  men- 
tioned. Successive  null  points  will  correspond 
to  the  indices  listed  in  the  table. 

A volume  indicator  in  the  transmitter  audio 
system  may  be  used  to  measure  the  audio  level 
required  for  different  amounts  of  deviation, 
and  the  indicator  thus  calibrated  in  terms  of 
frequency  deviation.  If  the  measurements  are 
made  at  the  fundamental  frequency  of  the  os- 
cillator, it  will  be  necessary  to  multiply  the 
frequency  deviation  by  the  harmonic  upon  which 
the  transmitter  is  operating,  of  course.  It  will 
probably  be  most  convenient  to  make  the  deter- 
mination at  some  frequency  intermediate  be- 
tween that  of  the  oscillator  and  that  at  which 
the  transmitter  is  operating,  and  then  to  mul- 
tiply the  result  by  the  frequency  multiplication 
between  that  frequency  and  the  transmitter 
output  frequency, 

16-4  Reception  of  FM 

Signals 


as  used  by  FM  broadcast  stations,  TV  sound, 
and  mobile  communications  FM. 

The  FM  receiver  must  have,  first  of  all,  a 
bandwidth  sufficient  to  pass  the  range  of  fre- 
quencies generated  by  the  FM  transmitter.  And 
since  the  receiver  must  be  a superheterodyne 
if  it  is  to  have  good  sensitivity  at  the  frequen- 
cies to  which  FM  is  restricted,  i-f  bandwidth 
is  an  important  factor  in  its  design. 

The  second  requirement  of  the  FM  receiver 
is  that  it  incorporate  some  sort  of  device  for 
converting  frequency  changes  into  amplitude 
changes,  in  other  words,  a detector  operating 
on  frequency  variations  rather  than  amplitude 
variations.  The  third  requirement,  and  one  which 
is  necessary  if  the  full  noise  reducing  capa- 


A conventional  communications  receiver  may 
be  used  to  receive  narrow-band  FM  transmis- 
sions, although  performance  will  be  much  poor- 
er than  can  be  obtained  with  an  NBFM  receiver 
or  adapter.  However,  a receiver  specifically 
designed  for  FM  reception  must  be  used  when 
it  is  desired  to  receive  high  deviation  FM  such 


Figure  13 

CATHODE-FOLLOWER  PHASE 
MODULATOR 

The  phase  modulator  illustrated  above  is 

quite  satisfactory  when  the  stage  is  to  he 
operated  on  a singie  frequency  or  over  a nar- 
row range  of  frequencies. 


322  FM  Transmission 


THE  RADIO 


Figure  14 

FM  RECEIVER  BLOCK  DIAGRAM 
Up  to  the  amplitude  limiter  stage,  the  FM 
receiver  is  similar  to  an  AM  receiver,  except 
for  a somewhat  wider  i~f  bandwidth.  The  lim, 
iter  removes  any  amplitude  modulation,  and 
the  frequency  detector  following  the  limiter 
converts  frequency  variations  into  amplitude 
variations. 


Figure  15 

SLOPE  DETECTION  OF  FM  SIGNAL 
One  side  of  the  response  characteristic  of  a 
tuned  circuit  or  of  an  i~f  amplifier  may  be 
used  as  shown  to  convert  frequency  varia- 
tions of  an  incoming  signal  into  amplitude 
variations. 


bilities  of  the  FM  system  of  transmission  are 
desired,  is  a limiting  device  to  eliminate  am- 
plitude variations  before  they  reach  the  de- 
tector. A block  diagram  of  the  essential  parts 
of  an  FM  receiver  is  shown  in  figure  14. 

The  Frequency  The  simplest  device  for  con- 

Detector  verting  frequency  variations 

to  amplitude  variations  is  an 
**off-tune”  resonant  circuit,  as  illustrated  in 
figure  15.  With  the  carrier  tuned  in  at  point 
"A,”  a certain  amount  of  t-f  voltage  will  be 
developed  across  the  tuned  circuit,  and,  as 
the  frequency  is  varied  either  side  of  this  fre- 
quency by  the  modulation,  the  r-f  voltage  will 
increase  and  decrease  to  points  "C”  and  "B” 
in  accordance  with  the  modulation.  If  the  volt- 
age across  the  tuned  circuit  is  applied  to  an 
ordinary  detector,  the  detector  output  will  vary 
in  accordance  with  the  modulation,  the  ampli- 
tude of  the  variation  being  proportional  to  the 
deviation  of  the  signal,  and  the  rate  being 
equal  to  the  modulation  frequency.  It  is  obvious 
from  figure  15  that  only  a small  portion  of  the 
resonance  curve  is  usable  for  linear  conversion 


Figure  16 

TRAVIS  DISCRIMINATOR 

This  type  of  discriminator  makes  use  of  two 
off-tuned  resonant  circuits  coupled  to  a sin- 
gle primary  winding.  The  circuit  is  capable 
of  excellent  linearity,  but  is  difficult  to  a- 
llgn. 


of  frequency  variations  into  amplitude  varia- 
tions, since  the  linear  portion  of  the  curve  is 
rather  short.  Any  frequency  variation  which 
exceeds  the  linear  portion  will  cause  distor- 
tion of  the  recovered  audio.  It  is  also  obvious 
by  inspection  of  figure  15  that  an  AM  receiver 
used  in  this  manner  is  wide  open  to  signals 
on  the  peak  of  the  resonance  curve  and  also 
to  signals  on  the  other  side  of  the  resonance 
curve.  Further,  no  noise  limiting  action  is  af- 
forded by  this  type  of  reception.  This  system, 
therefore,  is  not  recommended  for  FM  recep- 
tion, although  widely  used  by  amateurs  for 
occasional  NBFM  reception. 

Trovis  Discriminator  Another  form  of  frequen- 
cy detector  or  discrimi- 
nator, is  shown  in  figure  16.  In  this  arrange- 
ment two  tuned  circuits  are  used,  one  tuned 
on  each  side  of  the  i-f  amplifier  frequency, 
and  with  their  resonant  frequencies  spaced 
slightly  more  than  the  expected  transmitter 
swing.  Their  outputs  are  combined  in  a differ- 
ential rectifier  so  that  the  voltage  across  the 
series  load  resistors,  Rj  and  Rj,  is  equal  to 
the  algebraic  sum  of  the  individual  output 
voltages  of  each  rectifier.  When  a signal  at  the 


At  Its  "center"  fre- 
quency the  discrimi- 
nator produces  zero 
output  voltage.  On 
either  side  of  this 
frequency  It  gives 
a voltage  of  a polar- 
ity and  magnitude 
which  depend  on  the 
direction  and  amount 
of  frequency  shift. 

Figure  17 

DISCRIMINATOR  VOLTAGE-FREQUENCY 
CURVE 


FREQUENCY 


HANDBOOK 


FM  Reception  323 


This  discriminator  is  the  most  widely  used 
circuit  since  it  is  capoble  of  excellent  lin~ 
earity  and  is  relatively  simple  to  align  when 
proper  test  equipment  is  available. 


i-f  mid-frequency  is  received,  the  voltages 
across  the  load  resistors  are  equal  and  oppo- 
site, and  the  sum  voltage  is  zero.  As  the  t-f 
signal  varies  from  the  mid-frequency,  however, 
these  individual  voltages  become  unequal,  and 
a voltage  having  the  polarity  of  the  larger  volt- 
age and  equal  to  the  difference  between  the 
two  voltages  appears  across  the  series  resis- 
tors, and  is  applied  to  the  audio  amplifier. 
The  relationship  between  frequency  and  dis- 
criminator output  voltage  is  shown  in  figure 
17.  The  separation  of  the  discriminator  peaks 
and  the  linearity  of  the  output  voltage  vs.  fre- 
quency curve  depend  upon  the  discriminator 
frequency,  the  Q of  the  tuned  circuits,  and  the 
value  of  the  diode  load  resistors.  As  the  inter- 
mediate (and  discriminator)  frequency  is  in- 
creased, the  peaks  must  be  separated  further  to 
secure  good  linearity  and  output.  Within  limits, 
as  the  diode  load  resistance  or  the  Q is  re- 
duced, the  linearity  improves,  and  the  separa- 
tion between  the  peaks  must  be  greater. 

Foster-Seeley  The  most  widely  used  form  of 
Discriminator  discriminator  is  that  shown  in 
figure  18.  This  type  of  discrimi- 
nator yields  an  output-voltage-versus-frequen- 
cy  characteristic  similar  to  that  shown  in  fig- 
ure 19.  Here,  again,  the  output  voltage  is  equal 
to  the  algebraic  sum  of  the  voltages  developed 
across  the  load  resistors  of  the  two  diodes, 
the  resistors  being  connected  in  series  to 
ground.  However,  this  Foster-Seeley  discrim- 
inator requires  only  two  tuned  circuits  instead 
of  the  three  used  in  the  previous  discriminator. 
The  operation  of  the  circuit  results  from  the 
phase  relationships  existing  in  a transformer 
having  a tuned  secondary.  In  effect,  as  a close 
examination  of  the  circuit  will  reveal,  the  pri- 
mary circuit  is  in  series,  for  r.f.,  with  each 
half  of  the  secondary  to  ground.  When  the  re- 
ceived signal  is  at  the  resonant  frequency  of 
the  secondary,  the  r-f  voltage  across  the  sec- 
ondary is  90  degrees  out  of  phase  with  that 
across  the  primary.  Since  each  diode  is  con- 
nected across  one  half  of  the  secondary  wind- 


Figure  19 

DISCRIMINATOR  VECTOR  DIAGRAM 
A signal  at  the  resonant  frequency  of  the 
secondary  will  cause  the  secondary  voltage 
to  be  90  degrees  out  of  phase  with  the  pri- 
mary voltage,  as  shown  at  A,  and  the  result- 
ant voltages  R and  R'  are  equal.  If  the  sig- 
nal frequency  changes,  the  phase  relation- 
ship also  changes,  and  the  resultant  voltages 
are  no  longer  equal,  as  shown  at  B.  A differ- 
ential rectifier  is  used  to  give  an  output  volt- 
age proportional  to  the  difference  between 
R and  R'. 


ing  and  the  primary  winding  in  series,  the  re- 
sultant r-f  voltages  applied  to  each  are  equal, 
and  the  voltages  developed  across  each  diode 
load  resistor  are  equal  and  of  opposite  polar- 
ity. Hence,  the  net  voltage  between  the  top  of 
the  load  resistors  and  ground  is  zero.  This  is 
shown  vectorially  in  figure  19A  where  the  re- 
sultant voltages  R and  R ' which  ate  applied 
to  the  two  diodes  are  shown  to  be  equal  when 
the  phase  angle  between  primary  and  second- 
ary voltages  is  90  degrees.  If,  however,  the 
signal  varies  from  the  resonant  frequency,  the 
90-degree  phase  relationship  no  longer  exists 
between  primary  and  secondary.  The  result  of 
this  effect  is  shown  in  figure  19B  where  the 
secondary  r-f  voltage  is  no  longer  90  degrees 
out  of  phase  with  respect  to  the  primary  volt- 
age. The  resultant  voltages  applied  to  the  two 
diodes  are  now  no  longer  equal,  and  a d-c 
voltage  proportional  to  the  difference  between 
the  r-f  voltages  applied  to  the  two  diodes  will 
exist  across  the  series  load  resistors.  As  the 
signal  frequency  varies  back  and  forth  across 
the  resonant  frequency  of  the  discriminator, 
an  a-c  voltage  of  the  same  frequency  as  the 
original  modulation,  and  proportional  to  the 
deviation,  is  developed  and  passed  on  to  the 
audio  amplifier. 

Ratio  One  of  the  more  recent  types  of  FM 
Detector  detector  circuits,  called  the  ratio 
detector  is  diagrammed  in  figure  20. 
The  input  transformer  can  be  designed  so  that 
the  parallel  input  voltage  to  the  diodes  can  be 
taken  from  a tap  on  the  primary  of  the  trans- 


324  FM  Transmission 


THE  RADIO 


.0001  RFC 


The  parallel  voltage  to  the  diodes  in  a ratio 
detector  may  be  obtained  from  a tap  on  the 
primary  winding  of  the  transformer  or  from 
a third  winding.  Note  that  one  of  the  diodes 
is  reversed  from  the  system  used  with  the 
Foster-Seeley  discriminator^  and  that  the 
output  circuit  is  completely  different.  The 
ratio  detector  does  not  have  to  be  preceded 
by  a limiter,  but  is  more  difficult  to  align  for 
distortion-free  output  than  the  conventional 
discriminator. 


former,  or  this  voltage  may  be  obtained  from  a 
tertiary  winding  coupled  to  the  primary.  The 
r-f  choke  used  must  have  high  impedance  at 
the  intermediate  frequency  used  in  the  receiver, 
although  this  choke  is  not  needed  if  the  trans- 
former has  a tertiary  winding. 

The  circuit  of  the  ratio  detector  appears 
very  similar  to  that  of  the  more  conventional 
discriminator  arrangement.  However,  it  will  be 
noted  that  the  two  diodes  in  the  ratio  detector 
are  poled  so  that  their  d-c  output  voltages  add, 
as  contrasted  to  the  Foster-Seeley  circuit 
wherein  the  diodes  are  poled  so  that  the  d-c 
output  voltages  buck  each  other.  At  the  center 
frequency  to  which  the  discriminator  trans- 
former is  tuned  the  voltage  appearing  at  the 
top  of  the  Tmegohm  potentiometer  will  be  one- 
half  the  d-c  voltage  appearing  at  the  a-v-c  out- 
put terminal— since  the  contribution  of  each 
diode  will  be  the  same.  However,  as  the  input 
frequency  varies  to  one  side  or  the  other  of 
the  tuned  value  (while  remaining  within  the 
pass  band  of  the  i-f  amplifier  feeding  the  de- 
tector) the  relative  contributions  of  the  two 
diodes  will  be  different.  The  voltage  appearing 
at  the  top  of  the  1-megohm  volume  control  will 
increase  for  frequency  deviations  in  one  direc- 
tion and  will  decrease  for  frequency  deviations 
in  the  other  direction  from  the  mean  or  tuned 
value  of  the  transformer.  The  audio  output  volt- 
age is  equal  to  the  ratio  of  the  relative  contri- 
butions of  the  two  diodes,  hence  the  name 
ratio  detector. 

The  ratio  detector  offers  several  advantages 
over  the  simple  discriminator  circuit.  The  cir- 
cuit does  not  requite  the  use  of  a limiter  pre- 
ceding the  detector  since  the  circuit  is  inher- 
ently insensitive  to  amplitude  modulation  on 


Figure  21 

LIMITER  CIRCUIT 

One,  or  sometimes  two,  limiter  stages  nor- 
mally precede  the  discriminator  so  that  a con- 
stant signal  level  will  be  fed  to  the  FM  de- 
tector. This  procedure  eliminates  amplitude 
varirrtions  in  the  signal  fed  to  the  discrimi- 
nator, so  that  it  will  respond  only  to  frequen- 
cy changes. 


an  incoming  signal.  This  factor  alone  means 
that  the  r-f  and  i-f  gain  ahead  of  the  detector 
can  be  much  less  than  the  conventional  dis- 
criminator for  the  same  overall  sensitivity. 
Further,  the  circuit  provides  a-v-c  voltage  for 
controlling  the  gain  of  the  preceding  t-f  and 
i-f  stages.  The  ratio  detector  is,  however,  sus- 
ceptible to  variations  in  the  amplitude  of  the 
incoming  signal  as  is  any  other  detector  cir- 
cuit except  the  discriminator  with  a limiter 
preceding  it,  so  that  a-v-c  should  be  used  on 
the  stages  preceding  the  detector. 

Limiters  The  limiter  of  an  FM  receiver  using 
a convenrional  discriminator  serves 
to  remove  amplitude  modulation  and  pass  on 
to  the  discriminator  a frequency  modulated 
signal  of  constant  amplitude;  a typical  circuit 
is  shown  in  figure  21.  The  limiter  tube  is  oper- 
ated as  an  i-f  stage  with  very  low  plate  volt- 
age and  with  grid  leak  bias,  so  rhat  it  over- 
loads quite  easily.  Up  to  a certain  point  the 
output  of  the  limiter  will  increase  with  an  in- 
crease in  signal.  Above  this  point,  however, 
the  limiter  becomes  overloaded,  and  further 
large  increases  in  signal  will  not  give  any  in- 
crease in  ourput.  To  operate  successfully,  the 
limiter  must  be  supplied  with  a large  amount 
of  signal,  so  that  the  amplitude  of  its  output 
will  not  change  for  rather  wide  variations  in 
amplitude  of  the  signal.  Noise,  which  causes 
little  frequency  modulation  but  much  ampli- 
tude modulation  of  the  received  signal,  is  vir- 
tually wiped  out  in  the  limiter. 

The  voltage  across  the  grid  resistor  varies 
with  the  amplitude  of  the  received  signal.  For 
this  reason,  conventional  amplitude  modulated 
signals  may  be  received  on  the  FM  receiver 
by  connecting  the  input  of  the  audio  amplifier 
to  the  top  of  this  resistor,  rather  than  to  the 
discriminator  output.  When  properly  filtered 


HANDBOOK 


NBFM  Adapter  325 


by  a simple  R-C  circuit,  the  voltage  across 
the  grid  resistor  may  also  be  used  as  a-v-c 
voltage  for  the  receiver.  When  the  limiter  is 
operating  properly,  a.v.c.  is  neither  necessary 
nor  desirable,  however,  for  FM  reception  alone. 

Receiver  Design  One  of  the  most  important 
Considerations  factors  in  the  design  of  an 
FM  receiver  is  the  frequency 
swing  which  it  is  intended  to  handle.  It  will 
be  apparent  from  figure  17  that  if  the  straight 
portion  of  the  discriminator  circuit  covers  a 
wider  range  of  frequencies  than  those  gener- 
ated by  the  transmitter,  the  audio  output  will 
be  reduced  from  the  maximum  value  of  which 
the  receiver  is  capable. 

In  this  respect,  the  term  ’’modulation  per- 
centage” is  more  applicable  to  the  FM  receiver 
than  it  is  to  the  transmitter,  since  the  modula- 
tion capability  of  the  communication  system 
is  limited  by  the  receiver  bandwidth  and  the 
discriminator  characteristic;  full  utilization  of 
the  linear  portion  of  the  characteristic  amounts, 
in  effect,  to  100  per  cent  modulation.  This 
means  that  some  sort  of  standard  must  be  agreed 
upon,  for  any  particular  type  of  communication, 
to  make  it  unnecessary  to  vary  the  transmitter 
swing  to  accommodate  different  receivers. 

Two  considerations  influence  the  receiver 
bandwidth  necessary  for  any  particular  type  of 
communication.  These  are  the  maximum  audio 
frequency  which  the  system  will  handle,  and 
the  deviation  ratio  which  will  be  employed. 
For  voice  communication,  the  maximum  audio 
frequency  is  more  or  less  fixed  at  3000  to  4000 
cycles.  In  the  matter  of  deviation  ratio,  how- 
ever, the  amount  of  noise  suppression  which 
the  FM  system  will  provide  is  influenced  by 
the  ratio  chosen,  since  the  improvement  in 
signal-to-noise  ratio  which  the  FM  system 
shows  over  amplitude  modulation  is  equiva- 
lent to  a constant  multiplied  by  the  deviation 
ratio.  This  assumes  that  the  signal  is  some- 
what stronger  than  the  noise  at  the  receiver, 
however,  as  the  advantages  of  wideband  FM 
in  regard  to  noise  suppression  disappear  when 
the  signal-to-noise  ratio  approaches  unity. 

On  the  other  hand,  a low  deviation  ratio  is 
more  satisfactory  for  strictly  communication 
work,  where  readability  at  low  signal-to-noise 
ratios  is  more  important  than  additional  noise 
suppression  when  the  signal  is  already  appre- 
ciably stronger  than  the  noise. 

As  mentioned  previously,  broadcast  FM  prac- 
tice is  to  use  a deviation  ratio  of  5.  When  this 
ratio  is  applied  to  a voice-communication  sys- 
tem, the  total  swing  becomes  30  to  40  kc.  With 
lower  deviation  ratios,  such  as  are  most  fre- 
quently used  for  voice  work,  the  swing  becomes 

proportionally  less,  until  at  a deviation  ratio 

of  1 the  swing  is  equal  to  twice  the  highest 
audio  frequency.  Actually,  however,  the  re- 


R 

.02 

FROM 

DISCRIMINATOR 

i ' j 

TO  AUDIO  GRID 

c; 

T ; 

f*  1 MEC, 

R=  220  K, 
R=  100  K. 

C=  340  XIJJF. 

C = 750  U.UF. 

R = 47  K,  C = 1600  JJJJF. 
R=  22  K,  C = 3400  ilUF, 


Figure  22 

75-MICROSECOND  DE-EMPHASIS 
CIRCUITS 

The  audio  signal  transmitted  by  FM  and  TV 
stations  has  received  high-frequency  pre- 
emphasis, so  that  a de-emphasis  circuit 
should  be  included  between  the  output  of  the 
FM  detector  and  the  input  of  the  audio 
system. 


ceivet  bandwidth  must  be  slightly  greater  than 
the  expected  transmitter  swing,  since  for  dis- 
tortionless reception  the  receiver  must  pass 
the  complete  band  of  energy  generated  by  the 
transmitter,  and  this  band  will  always  cover  a 
range  somewhat  wider  than  the  transmitter 
swing. 

Pre-Emphasis  Standards  in  FM  broadcast 
and  De-Emphasis  and  TV  sound  work  call  for 
the  pre-emphasis  of  all  au- 
dio modulating  frequencies  above  about  2000 
cycles,  with  a rising  slope  such  as  would  be 
produced  by  a 75-raicrosecond  RL  network. 
Thus  the  FM  receiver  should  include  a com- 
pensating de-emphasis  RC  network  with  a time 
constant  of  75  microseconds  so  that  the  over- 
all frequency  response  from  microphone  to 
loudspeaker  will  approach  linearity.  The  use 
of  pre-emphasis  and  de-emphasis  in  this  man- 
ner results  in  a considerable  improvement  in 
the  overall  signal-to-noise  ratio  of  an  FM  sys- 
tem. Appropriate  values  for  the  de-emphasis 
network,  for  different  values  of  circuit  imped- 
ance ate  given  in  figure  22. 

A NBFM  4S5-kc.  The  unit  diagrammed  in  figure 
Adapter  Unit  23  is  designed  to  provide 
NBFM  reception  when  attached 
to  any  communication  receiver  having  a 455-kc. 
i-f  amplifier.  Although  NBFM  can  be  received 
on  an  AM  receiver  by  tuning  the  receiver  to 
one  side  or  the  other  of  the  incoming  signal, 
a tremendous  improvement  in  signal-to-noise 
ratio  and  in  signal  to  amplitude  ratio  will  be 
obtained  by  the  use  of  a true  FM  detector  sys- 
tem. 

The  adapter  uses  two  tubes.  A 6AU6  is 

used  as  a limiter,  and  a 6AL5  as  a discrimi- 
nator. The  audio  level  is  approximately  10 


326  FM  Transmission 


Radio  Teletype 


6AU6  6AL5 


NOTE  i ALL  CAPAC!  TORS  IN  JJF  UNLESS  OTHERWISE  NOTED 
ALL  RESISTORS  O.S  WATT  UNLESS  OTHERWISE  NOTED 


Figure  23 

NBFM  ADAPTER  FOR  45S.KC.  I-F  SYSTEM 


volts  peak  for  the  maximum  deviation  which 
can  be  handled  by  a conventional  455'kc.  i-f 
system.  The  unit  may  be  tuned  by  placing  a 
high  resistance  d-c  voltmeter  across  Ri  and 
tuning  the  trimmers  of  the  i-f  transformer  for 
maximum  voltage  when  an  unmodulated  signal 
is  injected  into  the  i-f  strip  of  the  receiver. 
The  voltmeter  should  next  be  connected  across 
the  audio  output  terminal  of  the  discriminator. 
The  receiver  is  now  tuned  back  and  forth  a- 
cross  the  frequency  of  the  incoming  signal, 
and  the  movement  of  the  voltmeter  noted.  When 
the  receiver  is  exactly  tuned  on  the  signal  the 
voltmeter  reading  should  be  zero.  When  the  re- 
ceiver is  tuned  to  one  side  of  center,  the  volt- 
meter reading  should  increase  to  a maximum 
value  and  then  decrease  gradually  to  zero  as 
the  signal  is  tuned  out  of  the  passband  of  the 
receiver.  When  the  receiver  is  tuned  to  the 
other  side  of  the  signal  the  voltmeter  should 
increase  to  the  same  maximum  value  but  in 
the  opposite  direction  or  polarity,  and  then 
fall  to  zero  as  the  signal  is  tuned  out  of  the 
passband.  It  may  be  necessary  to  make  small 
adjustments  to  Ci  and  Cj  to  make  the  volt- 
meter read  zero  when  the  signal  is  tuned  in 
the  center  of  the  passband. 

16-5  Radio  Teletype 

The  teletype  machine  is  an  electric  type- 
writer that  is  stimulated  by  d.c.  pulses  origi- 
nated by  the  action  of  a second  machine.  The 
pulses  may  be  transmitted  from  one  machine 
to  another  by  wire,  or  by  a radio  signal.  When 
radio  transmission  is  used,  the  system  is 
termed  radio  teletype  (RTTY). 

The  d.c.  pulses  that  comprise  the  teletype 
signal  may  be  converted  into  three  basic  types 
of  emission  suitable  for  radio  transmission. 
These  are;  1—  Frequency  shift  keying  (FSK), 
designated  as  FI  emission;  2—  Hake-break 


keying  (MBK), designated  as  A1  emission,  and; 

3—  Audio  frequency  shift  keying  (AFSK), 
designated  as  F2  emission. 

Frequency  shift  keying  is  obtained  by  vary- 
ing the  transmitted  frequency  of  the  radio 
signal  a fixed  amount  (usually  850  cycles) 
during  the  keying  process.  The  shift  is  ac- 
complished in  discrete  intervals  designated 
mark  and  space.  Both  types  of  intervals  convey 
information  to  the  teletype  printer.  Make-break 
keying  is  analogous  to  simple  c-w  transmission 
in  that  the  radio  carrier  conveys  information 
by  changing  from  an  off  to  an  on  condition. 
Early  RTTY  circuits  employed  MBK  equipment, 
which  is  rapidly  becoming  obsolete  since  it 
is  inferior  to  the  frequency  shift  system. 

Audio  frequency  shift  keying  employs  a 
steady  radio  carrier  modulated  by  an  audio 
tone  that  is  shifted  in  frequency  according  to 
the  RTTY  pulses.  Other  forms  of  information 
transmission  may  be  employed  by  a RTTY 
system  which  also  encompass  the  translation 
of  RTTY  pulses  into  r-f  signals. 

Teletype  The  RTTY  code  consists  of 

Coding  the  26  letters  of  the  alphabet, 

the  space,  the  line  feed,  the 
carriage  return,  the  bell,  the  upper  case  shift, 
and  the  lower  case  shift;  making  a total  of  32 
coded  groups.  Numerals,  punctuation,  and 
symbols  may  be  taken  care  of  in  the  case  shift, 
since  all  transmitted  letters  are  capitals. 

The  FSK  system  normally  employs  the  hi  gher 
radio  frequency  as  the  mark,  and  the  lower 
frequency  as  the  space.  This  relationship 
holds  true  in  the  AFSK  system  also.  The  lower 
audio  frequency  (mark)  is  normally  2125  cycles 
and  the  higher  audio  tone  (space)  is  2975 
cycles,  giving  a frequency  difference  of  850 
cycles. 

The  Teletype  A simple  FSK  teletype  system 
System  may  be  added  to  any  c-w  trans- 

mitter. The  teletype  keyboard 
prints  the  keyed  letters  on  a tape,  and  at  the 
same  time  generates  the  electrical  code  group 
that  describes  the  letter.  The  d.c.  pulses  are 
impressed  upon  a distributor  unit  which  ar- 
ranges the  typing  and  spacing  pulses  in  proper 
sequence.  The  resulting  series  of  impulses 
ate  applied  to  the  transmitter  frequency  con- 
trol device,  which  may  be  a reactance  modu- 
lator, actuated  by  a polar  relay. 

The  received  signal  is  hetrodyned  against 
a beat  oscillator  to  provide  the  two  audio  tones 
which  ate  limited  in  amplitude  and  passed 
through  audio  filters  to  separate  them.  Recti- 
fication of  the  tones  permits  operation  of  a 
polar  relay  which  can  provide  d.c.  pulses 
suitable  for  operation  of  the  tele-typewriter. 


CHAPTER  SEVENTEEN 


Sideband  Transmission 


While  single-sideband  transmission  (SSB) 
has  attracted  significant  interest  on  amateur 
frequencies  only  in  the  past  few  years,  the  prin- 
ciples have  been  recognized  and  put  to  use  in 
various  commercial  applications  for  many  years. 
Expansion  of  single-sideband  for  both  com- 
mercial and  amateur  communication  has  await- 
ed the  development  of  economical  components 
possessing  the  required  characteristics  (such  as 
sharp  cutoff  filters  and  high  stability  crystals) 
demanded  by  SSB  techniques.  The  availability 
of  such  components  and  precision  test  equip- 
ment now  makes  possible  the  economical  test- 
ing, adjustment  and  use  of  SSB  equipment  on 
a wider  scale  than  before.  Many  of  the  seem- 
ingly insurmountable  obstacles  of  past  years 
no  longer  prevent  the  amateur  from  achieving 
the  advantages  of  SSB  for  his  class  of  oper- 
ation. 

17-1  Commercial 

Applications  of  SSB 

Before  discussion  of  amateur  SSB  equipment, 
it  is  helpful  to  review  some  of  the  commercial 
applications  of  SSB  in  an  effort  to  avoid  prob- 
lems that  ate  already  solved. 

The  first  and  only  large  scale  use  of  SSB 
has  been  for  multiplexing  additional  voice  cir- 
cuits on  long  distance  telephone  toll  wires. 
Carrier  systems  came  into  wide  use  during  the 
30 ’s,  accompanied  by  the  development  of  high 
Q toroids  and  copper  oxide  ring  modulators 
of  controlled  characteristics. 


The  problem  solved  by  the  carrier  system 
was  that  of  translating  the  300-3000  cycle 
voice  band  of  frequencies  to  a higher  frequen- 
cy (for  example,  40.3  to  43.0  kc.)  for  trans- 
mission on  the  toll  wires,  and  then  to  reverse 
the  translation  process  at  the  receiving  termi- 
nal. It  was  possible  in  some  short-haul  equip- 
ment to  amplitude  modulate  a 40  kilocycle 
carrier  with  the  voice  frequencies,  in  which 
case  the  resulting  signal  would  occupy  a band 
of  frequencies  between  37  and  43  kilocycles. 
Since  the  transmission  properties  of  wires  and 
cable  deteriorate  rapidly  with  increasing  fre- 
quency, most  systems  required  the  bandwidth 
conservation  characteristics  of  single-sideband 
transmission.  In  addition,  the  carrier  wave  was 
generally  suppressed  to  reduce  the  power 
handling  capability  of  the  repeater  amplifiers 
and  diode  modulators.  A substantial  body  of 
literature  on  the  components  and  circuit  tech- 
niques of  SSB  has  been  generated  by  the  large 
and  continuing  development  effort  to  produce 
economical  carrier  telephone  systems. 

The  use  of  SSB  for  overseas  radiotelephony 
has  been  practiced  for  several  years  though 
the  number  of  such  circuits  has  been  numeri- 
cally small.  However,  the  economic  value  of 
such  circuits  has  been  great  enough  to  war- 
rant elaborate  station  equipment.  It  is  from 
these  stations  that  the  impression  has  been  ob- 
tained that  SSB  is  too  complicated  for  all  but 
a corps  of  engineers  and  technicians  to  handle. 
Components  such  as  lattice  filters  with  40 
or  more  crystals  have  suggested  astronomical 
expense. 


327 


328  Sideband  Transmission 


THE  RADIO 


REPRESENTATION  OF  A 
CONVENTIONAL  AM  SIGNAL 


More  recently,  SSB  techniques  have  been 
used  to  multiplex  large  numbers  of  voice  chan- 
nels on  a microwave  radio  band  using  equip- 
ment principally  developed  for  telephone  car- 
rier applications.  It  should  be  noted  that  all 
production  equipment  employed  in  these  ser- 
vices uses  the  filter  method  of  generating  the 
single-sideband  signal,  though  there  is  a wide 
variation  in  the  types  of  filters  actually  used. 
The  SSB  signal  is  generated  at  a low  fre- 
quency and  at  a low  level,  and  then  trans- 
lated and  linearly  amplified  to  a high  level 
at  the  operating  frequency. 

Considerable  development  effort  has  been 
expended  on  high  level  phasing  type  trans- 
mitters wherein  the  problems  of  linear  ampli- 
fication are  exchanged  for  the  problems  of 
accurately  controlled  phase  shifts.  Such  equip- 
ment has  featured  automatic  tuning  circuits, 
servo-driven  to  facilitate  frequency  changing, 
but  no  transmitter  of  this  type  has  been  suffi- 
ciently attractive  to  warrant  appreciable  pro- 
duction. 

17-2  Derivation  of 
Single-Sideband  Signals 

The  single-sideband  method  of  communica- 
tion is,  essentially,  a procedure  for  obtaining 
more  efficient  use  of  available  frequency  spec- 
trum and  of  available  transmitter  capability. 
As  a starting  point  for  the  discussion  of  sin- 
gle-sideband signals,  let  us  take  a conven- 
tional AM  signal,  such  as  shown  in  figure  1, 
as  representing  the  most  common  method  for 
transmitting  complex  intelligence  such  as 
voice  or  music. 

It  will  be  noted  in  figure  1 that  there  are 
three  distinct  portions  to  the  signal:  the  car- 
rier, and  the  upper  and  the  lower  sideband 
group.  These  three  portions  always  are  present 
in  a conventional  AM  signal.  Of  all  these  por- 
tions the  carrier  is  the  least  necessary  and 
the  most  expensive  to  transmit.  It  is  an  actual 


fact,  and  it  can  be  proved  mathematically  (and 

physically  with  a highly  selective  receiver) 
that  the  carrier  of  an  AM  signal  remains  un- 
changed in  amplitude,  whether  it  is  being  mod- 
ulated or  not.  Of  course  the  carrier  appears 
to  be  modulated  when  we  observe  the  modu- 
lated signal  on  a receiving  system  or  indicator 
which  passes  a sufficiently  wide  band  that 
the  carrier  and  the  modulation  sidebands  are 
viewed  at  the  same  time.  This  apparent  change 
in  the  amplitude  of  the  carrier  with  modula- 
tion is  simply  the  result  of  the  sidebands  beat- 
ing with  the  carrier.  However,  if  we  receive 
the  signal  on  a highly  selective  receiver,  and  if 
we  modulate  the  carrier  with  a sine  wave  of 
3000  to  5000  cycles,  we  will  readily  see  that 
the  carrier,  or  either  of  the  sidebands  can  be 
tuned  in  separately;  the  carrier  amplimde,  as 
observed  on  a signal  strength  meter,  will  re- 
main constant,  while  the  amplitude  of  the  side- 
bands will  vary  in  direct  proportion  to  the 
modulation  percentage. 

Elimination  of  It  is  obvious  from  the  pre- 
the  Carrier  and  vious  discussion  that  the 
One  Sideband  carrier  is  superfluous  so  far 
as  the  transmission  of  in- 
telligence is  concerned.  It  is  obviously  a con- 
venience, however,  since  it  provides  a signal 
at  the  receiving  end  for  the  sidebands  to  beat 
with  and  thus  to  reproduce  the  original  mod- 
ulating signal.  It  is  equally  true  that  the  trans- 
mission of  both  sidebands  under  ordinary  con- 
ditions is  superfluous  since  identically  the 
same  intelligence  is  contained  in  both  side- 
bands. Several  systems  for  carrier  and  side- 
band elimination  will  be  discussed  in  this 
chapter. 

Power  Advantage  Single  sideband  is  a very 
of  SSB  over  AM  efficient  form  of  voice 
communication  by  radio. 
The  amount  of  radio  frequency  spectrum  oc- 
cupied can  be  no  greater  than  the  frequency 
range  of  the  audio  or  speech  signal  transmitted, 
whereas  other  forms  of  radio  transmission  re- 
quire from  two  to  several  times  as  much  spec- 
trum space.  The  r-f  power  in  the  transmitted 
SSB  signal  is  directly  proportional  to  the  power 
in  the  original  audio  signal  and  no  strong 
carrier  is  transmitted.  Except  for  a weak  pilot 
carrier  present  in  some  commercial  usage, 
there  is  no  r-f  output  when  there  is  no  audio 
input. 

The  power  output  tating  of  a SSB  trans- 
mitter is  given  in  terms  of  peak  envelope 
power  (PEP).  This  may  be  defined  as  the 
r-m-s  power  at  the  crest  of  the  modulation 


HANDBOOK 


Derivation  329 


envelope.  The  peak  envelope  power  of  a con- 
ventional amplitude  modulated  signal  at  100% 
modulation  is  four  times  the  carrier  power. 
The  average  power  input  to  a SSB  transmitter 
is  therefore  a very  small  fraction  of  the  power 
input  to  a conventional  amplitude  modulated 
transmitter  of  the  same  power  rating. 

Single  sideband  is  well  suited  for  long- 
range  communications  because  of  its  spectrum 
and  power  economy  and  because  it  is  less  sus- 
ceptible to  the  effects  of  selective  fading  and 
interference  than  amplitude  modulation.  The 
principal  advantages  of  SSB  arise  from  the 
elimination  of  the  high-energy  carrier  and 
from  further  reduction  in  sideband  power  per- 
mitted by  the  improved  performance  of  SSB 
under  unfavorable  propagation  conditions. 

In  the  presence  of  narrow  band  man-made 
interference,  the  narrower  bandwidth  of  SSB 
reduces  the  probability  of  destructive  inter- 
ference .A  statistical  study  of  the  distribution 
of  signals  on  the  ait  versus  the  signal  strength 
shows  that  the  probability  of  successful  com- 
munication will  be  the  same  if  the  SSB  power 
is  equal  to  one-half  the  power  of  one  of  the 
two  a-m  sidebands.  Thus  SSB  can  give  from  0 
to  9 db  improvement  under  various  conditions 
when  the  total  sideband  power  is  equal  in 
SSB  and  a-m.  In  general,  it  may  be  assumed 
that  3 db  of  the  possible  9 db  advantage  will 
be  realized  on  the  average  contact.  In  this  case, 
the  SSB  power  required  for  equivalent  per- 
formance is  equal  to  the  power  in  one  of  the 
a-m  sidebands.  For  example,  this  would  rate 
a 100-watt  SSB  and  a 400  watt  (carrier)  a-m 
transmitter  as  having  equal  performance.  It 
should  be  noted  that  in  this  comparison  it  is 
assumed  that  the  receiver  bandwidth  is  just 
sufficient  to  accept  the  transmitted  intelligence 
in  each  case. 

To  help  evaluate  other  methods  of  compari- 
son the  following  points  should  be  considered. 
In  conventional  amplitude  modulation  two 
sidebands  are  transmitted,  each  having  a peak 
envelope  power  equal  to  14 -carrier  power.  For 
example,  a 100-watt  a-m  signal  will  have  25- 
watt  peak  envelope  power  in  each  sideband, 
or  a total  of  50  watts.  When  the  receiver  de- 
tects this  signal,  the  voltages  of  the  two  side- 
bands are  added  in  the  detector.  Thus  the  de- 
tector output  voltage  is  equivalent  to  that  of 
a 100-watt  SSB  signal.  This  method  of  com- 
parison says  that  a 100  watt  SSB  transmitter 
is  just  equivalent  to  a 100-watt  a-m  trans- 
mitter. This  assumption  is  valid  only  when  the 

receiver  bandwidth  used  for  SSB  is  the  same 
as  that  requited  for  amplitude  modulation 


^ATt 


'4000  KC,  4004  KC^ 


AUDIO  SPECTRUM 
@ 


SSB  SPECTRUM 
{UPPER  SIDEBAND) 


SSB  SPECTRUM 
{LOWER  SIDEBAND ) 

© 


Figure  2 

RELATIONSHIP  OF  AUDIO  AND 
SSB  SPECTRUMS 


T.'te  single  sideband  components  are  the  same 
as  the  original  audio  components  except  that 
the  frequency  of  each  is  raised  by  the  fre- 
quency of  the  carrier.  The  relative  amplitude 
of  the  various  components  remains  the  same. 


(e.g.,  6 kilocycles),  when  there  is  no  noise  or 
interference  other  than  broadband  noise,  and 
if  the  a-m  signal  is  not  degraded  by  propaga- 
tion. By  using  half  the  bandwidth  for  SSB 
reception  (e.g.,  3 kilocycles)  the  noise  is  re- 
duced 3 db  so  the  100  watt  SSB  signal  be- 
comes equivalent  to  a 200  watt  carrier  a-m 
signal.  It  is  also  possible  for  the  a-m  signal 
to  be  degraded  another  3 db  on  the  average 
due  to  narrow  band  interference  and  poor 
propagation  conditions,  giving  a possible  4 
to  1 power  advantage  to  the  SSB  signal. 

It  should  be  noted  that  3 db  signal-to-noise 
ratio  is  lost  when  receiving  only  one  sideband 
of  an  a-m  signal.  The  narrower  receiving  band- 
width reduces  the  noise  by  3 db  but  the  6 db 
advantage  of  coherent  detection  is  lost,  leaving 
a net  loss  of  3 db.  Poor  propagation  will  de- 
grade this  "one  sideband”  reception  of  an  a-m 
signal  less  than  double  sideband  reception, 
however.  Also  under  severe  narrow  band  in- 
terference conditions  (e.g.,  an  adjacent  strong 
signal ) the  ability  to  reject  all  interference  on 
one  side  of  the  carrier  is  a great  advantage. 

The  Noture  of  a The  nature  of  a single 
SSB  Signal  sideband  signal  is  easily 

visualized  by  noting  that 
the  SSB  signal  components  are  exactly  the  same 
as  the  original  audio  components  except  that 
the  frequency  of  each  is  raised  by  the  frequen- 
cy of  the  carrier.  The  relative  amplimde  of  the 
various  components  remains  the  same,  how- 
ever. (The  first  statement  is  only  true  for  the 
upper  sideband  since  the  lower  sideband  fre- 
quency components  are  the  difference  between 
the  carrier  and  the  original  audio  signal ) . 
Figure  2A,  B,  and  C shows  how  the  audio 
spectrum  is  simply  moved  up  into  the  radio 
spectrum  to  give  the  upper  sideband.  The 

lower  sideband  is  the  same  except  inverted,  as 
shown  in  figure  2C.  Either  sideband  may  be 
used.  It  is  apparent  that  the  carrier  frequency 


330  Sideband  Transmission 


THE  RADIO 


Figure  3 

A SINGLE  SINE  WAVE  TONE  INPUT 
TO  A SSB  TRANSMITTER  RESULTS 
IN  A STEADY  SINGLE  SINE  WAVE 
R-F  OUTFIT  (A).  TWO  AUDIO  TONES 
OF  EQUAL  AMPLITUDE  BEAT 
TOGETHER  TO  PRODUCE  HALF-SINE 
WAVES  AS  SHOWN  IN  (B). 


of  a SSB  signal  can  only  be  changed  by  add- 
ing or  subtracting  to  the  original  carrier  fre- 
quency. This  is  done  by  heterodyning,  using 
converter  or  mixer  circuits  similar  to  those 
employed  in  a superheterodyne  receiver. 

It  is  noted  that  a single  sine  wave  tone  in- 
put to  a SSB  transmitter  results  in  a single 
steady  sine  wave  r-f  ouput,  as  shown  in  figure 
3A.  Since  it  is  difficult  to  measure  the  per- 
formance of  a linear  amplifier  with  a single 
tone,  it  has  become  stand  ird  practice  to  use 
two  tones  of  equal  amplitude  for  test  pur- 
poses. The  two  radio  frequencies  thus  pro- 
duced beat  together  to  give  the  SSB  envelope 
shown  in  figure  3B.  This  figure  has  the  shape 
of  half  sine  waves,  and  from  one  null  to  the 
next  represents  one  full  cycle  of  the  difference 
frequency.  How  this  envelope  is  generated  is 
shown  more  fully  in  figures  4A  and  4B.  fi 
and  B represent  the  two  tone  signals.  When 
a vector  representing  the  lower  frequency  tone 
signal  is  used  as  a reference,  the  other  vector 
rotates  around  it  as  shown,  and  this  action 


FREQUENCY  f'i 


© 


Figure  4 

VECTOR  REPRESENTATION  OF 
TWO-TONE  SSB  ENVELOPE 


Figure  5 
TWO-TONE  SSB 
ENVELOPE  WHEN 
ONE  TONE  HAS 
TWICE  THE 
AMPLITUDE  OF 
THE  OTHER. 


Figure  6 

THREE-TONE  SSB 
ENVELOPE  WHEN 
EQUAL  TONES  OF 
EQUAL  FREQUENCY 
SPACINGS 
ARE  USED. 


generates  the  SSB  envelope  When  the  two 
vectors  are  exactly  opposite  in  phase,  the  out- 
put is  zero  and  this  causes  the  null  in  the  en- 
velope. If  one  tone  has  twice  the  amplitude  of 
the  other,  the  envelope  shape  is  shown  in 
figure  5.  Figure  6 shows  the  SSB  envelope  of 
three  equal  tones  of  equal  frequency  spacings 
and  at  one  particular  phase  relationship.  Figure 
7 A shows  the  SSB  envelope  of  four  equal 
tones  with  equal  frequency  spacings  and  at 
one  particular  phase  relationship.  The  phase 
relationships  chosen  are  such  that  at  some  in- 
stant the  vectors  representing  the  several  tones 
are  all  in  phase.  Figure  7B  shows  a SSB  envel- 
ope of  a square  wave.  A pure  square  wave  re- 
quires infinite  bandwidth,  so  its  SSB  envelope 
requires  infinite  amplitude.  This  emphasizes 
the  point  that  the  SSB  envelope  shape  is  not 
the  same  as  the  original  audio  wave  shape,  and 
usually  bears  no  similarity  to  it.  This  is  be- 
cause the  percentage  difference  between  the 
radio  frequencies  is  very  small,  even  though 
one  audio  tone  may  be  several  times  the  other 
in  terms  of  frequency.  Speech  clipping  as  used 


Figure  7A 
FOUR  TONE 
SSB  ENVELOPE 

when  equal  tones 
with  equal  frequency 
spacings  are  used 


Figure  7B 
SSB  ENVELOPE 
OF  A SQUARE 
WAVE. 

Peak  of  wave  reaches 
infinite  amplitude. 


HANDBOOK 


Derivation  331 


in  amplitude  modulation  is  of  no  practical 
value  in  SSB  because  the  SSB  r-f  envelopes 
are  so  different  than  the  audio  envelopes.  A 
heavily  clipped  wave  approaches  a square  wave 
and  a square  wave  gives  a SSB  envelope  with 
peaks  of  infinite  amplitude  as  shown  in  figure 
7B. 

Carrier  Frequency  Reception  of  a SSB 

Stability  Requirements  signal  is  accom- 

plished by  simply 
heterodyning  the  carrier  down  to  zero  fre- 
quency. (The  conversion  frequency  used  in  the 
last  heterodyne  step  is  often  called  the  rein- 
serted. carrier).  If  the  SSB  signal  is  not  hetero- 
dyned down  to  exactly  zero  frequency,  each 
frequency  component  of  the  detected  audio 
signal  will  be  high  or  low  by  the  amount  of 
this  error.  An  error  of  10  to  20  c.p  s.  for  speech 
signals  is  acceptable  from  an  intelligibility 
standpoint,  but  an  error  of  the  order  of  50 
c.p.s.  seriously  degrades  the  intelligibility.  An 
error  of  20  c.p.s.  is  not  acceptable  for  the 
transmission  of  music,  however,  because  the 
harmonic  relationship  of  the  notes  would  be 
destroyed.  For  example,  the  harmonics  of  220 
c.p.s.  are  440,  660,  880,  etc.,  but  a 10  c.p  s. 
error  gives  230,  450,  670,  890,  etc.,  or  210, 
430,  650,  870,  etc.,  if  the  original  error  is  on 
the  other  side.  This  error  would  destroy  the 
original  sound  of  the  tones,  and  the  harmony 
between  the  tones. 

Suppression  of  the  carrier  is  common  in  ama- 
teur SSB  work,  so  the  combined  frequency  sta- 
bilities of  all  oscillators  in  both  the  transmit- 
ting and  receiving  equipment  add  together  to 
give  the  frequency  error  found  in  detection. 
In  order  to  overcome  much  of  the  frequency 
stability  problem,  it  is  common  commercial 
practice  to  transmit  a pilot  carrier  at  a re- 
duced amplitude.  This  is  usually  20  db  below 
one  tone  of  a two-tone  signal,  or  26  db  below 
the  peak  envelope  power  rating  of  the  trans- 
mitter. This  pilot  carrier  is  filtered  out  from 
the  other  signals  at  the  receiver  and  either  am- 
plified and  used  for  the  reinserted  carrier  or 
used  to  control  the  frequency  of  a local  oscil- 
lator. By  this  means,  the  frequency  drift  of 
the  carrier  is  eliminated  as  an  error  in  detec- 
tion. 

Advantage  of  SSB  On  long  distance  com- 

with  Selective  Fading  munication  circuits 
using  a-m,  selective 
fading  often  causes  severe  distortion  and  at 
times  makes  the  signal  unintelligible.  When 
one  sideband  is  weaker  than  the  other,  distor- 


Figure  8 

SHOWING  TWO  COMMON  TYPES 
OF  BALANCED  MODULATORS 

Notice  that  a balanced  modulator  changes 
the  circuit  condition  from  single  ended  to 
push-pull,  or  vice  versa.  Choice  of  circuit  de- 
pends upon  external  circuit  conditions  since 
both  the  (A)  and  (B)  arrangements  can  give 
satisfactory  generation  of  a double-sideband 
suppressed-carrier  signal. 


tion  results;  but  when  the  carrier  becomes 
weak  and  the  sidebands  are  strong,  the  distor- 
tion is  extremely  severe  and  the  signal  may 
sound  like  "monkey  chatter.”  This  is  because 
a carrier  of  at  least  twice  the  amplitude  of 
either  sideband  is  necessary  to  demodulate  the 
signal  properly.  This  can  be  overcome  by  us- 
ing exalted  carrier  reception  in  which  the  car- 
rier is  amplified  separately  and  then  reinserted 
before  the  signal  is  demodulated  or  detected. 
This  is  a great  help,  but  the  reinserted  carrier 
must  be  very  close  to  the  same  phase  as  the 
original  carrier.  For  example,  if  the  reinserted 
carrier  were  90  degrees  from  the  original 
source,  the  a-m  signal  would  be  converted  to 
phase  modulation  and  the  usual  a-m  detector 
would  deliver  no  output. 

The  phase  of  the  reinserted  carrier  is  of  no 
importance  in  SSB  reception  and  by  using  a 
strong  reinserted  carrier,  exalted  carrier  recep- 
tion is  in  effect  realized.  Selective  fading  with 
one  sideband  simply  changes  the  amplitude 
and  the  frequency  response  of  the  system  and 
very  seldom  causes  the  signal  to  become  unin- 
telligible. Thus  the  receiving  techniques  used 
with  SSB  ate  those  which  inherently  greatly 
minimize  distortion  due  to  selective  fading. 


332  Sideband  Transmission 


THE  RADIO 


VOLTAGE 

Figure  9 

TWO  TYPES  OF  DIODE  BALANCED 
MODULATOR 

Such  balanced  modulator  circuits  are  com* 
monly  used  in  carrier  telephone  work  and  in 
single-sideband  systems  where  the  carrier 
frequency  and  modulating  frequency  are  rela- 
tively close  together.  Vacuum  diodes,  copper- 
oxide  rectifiers,  or  crystal  diodes  may  be 
used  in  the  circuits. 


17-3  Carrier  Eliminafion 
Circuits 

Various  circuits  may  be  employed  to  elimi- 
nate the  carrier  to  provide  a double  sideband 
signal.  A selective  filter  may  follow  the  carrier 
elimination  circuit  to  produce  a single  side- 
band signal. 

Two  modulated  amplifiers  may  be  connected 
with  the  carrier  inputs  180°  out  of  phase,  and 
with  the  carrier  outputs  in  parallel.  The  car- 


rier will  be  balanced  out  of  the  output  circuit, 
leaving  only  the  two  sidebands.  Such  a cir- 
cuit is  called  a balanced  modulator. 

Any  non-linear  element  will  produce  modu- 
lation. That  is,  if  two  signals  are  put  in,  sum 
and  difference  frequencies  as  well  as  the  orig- 
inal frequencies  appear  in  the  output.  This 
phenomenon  is  objectionable  in  amplifiers  and 
desirable  in  modulators  or  mixers. 

In  addition  to  the  sum  and  difference  fre- 
quencies, other  outputs  ( such  as  twice  one 
frequency  plus  the  other)  may  appear.  All 
combinations  of  all  harmonics  of  each  input 
frequency  may  appear,  but  in  general  these  are 
of  decreasing  amplitude  with  increasing  order 
of  harmonic.  These  outputs  are  usually  re- 
jected by  selective  circuits  following  the  mod- 
ulator. All  modulators  are  not  alike  in  the 
magnitude  of  these  higher  order  outputs.  Bal- 
anced diode  rings  operating  in  the  square  law 
region  are  fairly  good  and  pentagrid  converters 
much  poorer.  Excessive  carrier  level  in  tube 
mixers  will  increase  the  relative  magnitude 
of  the  higher  order  outputs.  Two  types  of 
triode  balanced  modulators  are  shown  in  figure 

8,  and  two  types  of  diode  modulators  in  figure 

9.  Balanced  modulators  employing  vacuum 
tubes  may  be  made  to  work  very  easily  to  a 
point.  Circuits  may  be  devised  wherein  both 
input  signals  may  be  applied  to  a high  im- 
pedance grid,  simplifying  isolation  and  load- 
ing problems.  The  most  important  difficulties 
with  these  vacuum  tube  modulator  circuits 
are;  ( 1 ) Balance  is  not  independent  of  signal 
level.  (2)  Balance  drifts  with  time  and  envir- 
onment. ( 3 ) The  carrier  level  for  low  "high- 
order  output”  is  critical,  and  (4)  Such  circuits 
have  limited  dynamic  range. 

A number  of  typical  circuits  are  shown  in 
figure  10.  Of  the  group  the  most  satisfactory 
performance  is  to  be  had  from  plate  modulated 
triodes. 


PUSH-PULL  BT 

AUDIO  IN 


@ 

PLATE  MODULATED  BALANCED 
TRIODE  MODULATOR 


BALANCED  TRIODE  MODULATOR  BALANCED  PENTAGRID  CON- 

WITH  SINGLE  ENDED  INPUT  CIRCUITS  VERTER  MODULATOR 


Figure  10 

BALANCED  MODULATORS 


HANDBOOK 


Carrier  Elimination  333 


@ 


DOUBLE-BALANCED  RING  MODULATOR 


SHUNT- QUAD  MODULATOR 

Figure  1 1 

DIODE  RING  MODULATORS 


Diode  Ring  Modulation  in  telephone  car- 
Modulators  tier  equipment  has  been  very 
successfully  accomplished  with 
copper-oxide  double  balanced  ring  modulators. 
More  recently,  germanium  diodes  have  been 
applied  to  similar  circuits.  The  basic  diode 
ring  circuits  are  shown  in  figure  11.  The  most 
widely  applied  is  the  double  balanced  ring 
(A).  Both  carrier  and  input  are  balanced  with 
respect  to  the  output,  which  is  advantageous 
when  the  output  frequency  is  not  sufficiently 
different  from  the  inputs  to  allow  ready  sep- 
aration by  filters.  It  should  be  noted  that  the 
carrier  must  pass  through  the  balanced  input 
and  output  transformers.  Care  must  be  taken 
in  adapting  this  circuit  to  minimize  the  carrier 
power  that  will  be  lost  in  these  elements.  The 
shunt  and  series  quad  circuits  are  usable  when 
the  output  frequencies  are  entirely  different 
(i.e.:  audio  and  r.f. ).  The  shunt  quad  (B) 
is  used  with  high  source  and  load  impedances 
and  the  series  quad  (C)  with  low  source  and 
load  impedances.  These  two  circuits  may  be 
adapted  to  use  only  two  diodes,  substituting  a 
balanced  transformer  for  one  side  of  the 
bridge,  as  shown  in  figure  12.  It  should  be 
noted  that  these  circuits  present  a half-wave 
load  to  the  carrier  source.  In  applying  any  of 
these  circuits,  r-f  chokes  and  capacitors  must 
be  employed  to  control  the  path  of  signal  and 
carrier  currents.  In  the  shunt  pair,  for  example, 
a blocking  capacitor  is  used  to  prevent  the  r-f 
load  from  shorting  the  audio  input. 

To  a first  approximation,  the  source  and 
load  impedances  should  be  an  arithmetical 
mean  of  the  forward  and  back  resistances  of 
the  diodes  employed.  A workable  rule  of 
thumb  is  that  the  source  and  load  impedances 
be  ten  to  twenty  times  the  forward  resistance 
for  semi-conductor  rings.  The  high  frequency 
limit  of  operation  in  the  case  of  junction  and 
copper-oxide  diodes  may  be  appreciably  ex- 
tended by  the  use  of  very  low  source  and  load 
impedances. 

Copper-oxide  diodes  suitable  for  carrier 


work  are  normally  manufactured  to  order.  They 
offer  no  particular  advantage  to  the  amateur, 
though  their  excellent  long-term  stability  is 
important  in  commercial  applications.  Recti- 
fier types  intended  to  be  used  as  meter  recti- 
fiers are  not  likely  to  have  the  balance  or  high 
frequency  response  desirable  in  amateur  SSB 
transmitters. 

Vacuum  diodes  such  as  the  6AL5  may  be 
used  as  modulators.  Balancing  the  heater- 
cathode  capacity  is  a major  difficulty  except 
when  the  6AL5  is  used  at  low  source  and  load 
impedance  levels.  In  addition,  contact  poten- 
tials of  the  order  of  a few  tenths  of  a volt  may 
also  disturb  low  level  applications  (figure  13). 

The  double  diode  circuits  appear  attractive, 
but  in  general  it  is  more  difficult  to  balance  a 
transformer  at  carrier  frequency  than  an  addi- 
tional pair  of  diodes.  Balancing  potentiometers 
may  be  employed,  but  the  actual  cause  of  the 
unbalance  is  far  more  subtile,  and  cannot  be 
adequately  corrected  with  a single  adjustment. 

A signal  produced  by  any  of  the  above  cir- 
cuits may  be  classified  as  a double  sideband, 
suppressed-carrier  signal. 


@ 

SHUNT-PAIR 

MODULATOR 


® 

SERIES-PAIR 

MODULATOR 


Figure  12 

DOUBLE-DIODE  PAIRED  MODULATORS 


334  Sideband  Transmission 


THE  RADIO 


(B)  ring-diode  modulator  using  6AL5  TUBE 

Figure  13 

VACUUM  DIODE  MODULATOR  CIRCUITS 

17-4  Generation  of 

Single-Sideband  Signals 

In  general,  there  are  two  commonly  used 
methods  by  which  a single-sideband  signal  may 
be  generated.  These  systems  are:  (1)  The  Fil- 
ter Method,  and  (2)  The  Phasing  Method. 
The  systems  may  be  used  singly  or  in  com- 
bination, and  either  method,  in  theory,  may 
be  used  at  the  operating  frequency  of  the 
transmitter  or  at  some  other  frequency  with 
the  signal  at  the  operating  frequency  being  ob- 
tained through  the  use  of  frequency  changers 
( mixers ) . 

The  Filter  The  filter  method  for  obtaining 
Method  a SSB  signal  is  the  classic  meth- 
od which  has  been  in  use  by  the 
telephone  companies  for  many  years  both  for 


Figure  15 

BANDPASS  CHARACTERISTIC  OF 
BURNELL  S-1S000  SINGLE 
SIDEBAND  FILTER 


land-line  and  radio  communications.  The  mode 
of  operation  of  the  filter  method  is  diagram- 
med in  figure  14,  in  terms  of  components  and 
filters  which  normally  would  be  available  to 
the  amateur  or  experimenter.  The  output  of 
the  speech  amplifier  passes  through  a con- 
ventional speech  filter  to  limit  the  frequency 
range  of  the  speech  to  about  200  to  3000 
cycles.  This  signal  then  is  fed  to  a balanced 
modulator  along  with  a 50,000-cycle  first  car- 
rier from  a self-excited  oscillator.  A low-fre- 
quency balanced  modulator  of  this  type  most 
conveniently  may  be  made  up  of  four  diodes 
of  the  vacuum  or  crystal  type  cross  connected 
in  a balanced  bridge  or  ring  modulator  circuit. 
Such  a modulator  passes  only  the  sideband 
components  resulting  from  the  sum  and  dif- 
ference between  the  two  signals  being  fed  to 
the  balanced  modulator.  The  audio  signal  and 
the  50-kc.  carrier  signal  from  the  oscillator 
both  cancel  out  in  the  balanced  modulator  so 
that  a band  of  frequencies  between  47  and  50 
kc.  and  another  band  of  frequencies  between 
50  and  53  kc.  appear  in  the  output. 

The  signals  from  the  first  balanced  modu- 
lator are  then  fed  through  the  most  critical 


foo-roooo'b  £00~3000'\y  47-SOkC.  47~30KC.  47-5QKC. 


Figure  14 

BLOCK  DIAGRAM  OF  FILTER  EXCITER  EMPLOYING  A 50-K.C. 
SIDEBAND  FILTER 


HANDBOOK 


Generation  of  S.S.B.  335 


6AU6  -^laAU?  6AL5  ^ I2AU7  6C4 


50-KC.  SIDEBAND  FILTER 


component  in  the  whole  system — the  first  side- 
band filter.  It  is  the  function  of  this  first  side- 
band filter  to  separate  the  desired  47  to  50  kc. 
sideband  from  the  unneeded  and  undesired  50 
to  53  kc.  sideband.  Hence  this  filter  must  have 
low  attenuation  in  the  region  between  47  and 
50  kc.,  a very  rapid  slope  in  the  vicinity  of 
50  kc.,  and  a very  high  attenuation  to  the 
sideband  components  falling  between  50  and 
53  kilocycles. 

Burnell  & Co.,  Inc.,  of  Yonkers,  New  York 
produce  such  a filter,  designated  as  Burnell 
S- 15,000.  The  passband  of  this  filter  is  shown 
in  figure  15- 

Appearing,  then,  at  the  output  of  the  filter 
is  a single  sideband  of  47  kc.  to  50  kc.  This 
sideband  may  be  passed  through  a phase  in- 
verter to  obtain  a balanced  output,  and  then 
fed  to  a balanced  mixer.  A local  oscillator 
operating  in  the  range  of  1750  kc.  to  1950  kc. 
is  used  as  the  conversion  oscillator.  Additional 
conversion  stages  may  now  be  added  to  trans- 


late the  SSB  signal  to  the  desired  frequency. 
Since  only  linear  amplification  may  be  used, 
it  is  not  possible  to  use  frequency  multiplying 
stages.  Any  frequency  changing  must  be  done 
by  the  beating-oscillator  technique.  An  oper- 
ational circuit  of  this  type  of  SSB  exciter  is 
shown  in  figure  16. 

A second  type  of  filter-exciter  for  SSB  may 
be  built  around  the  Collins  Mechanical  Filter. 
Such  an  exciter  is  diagrammed  in  figure  17. 
Voice  frequencies  in  the  range  of  200-3000 
cycles  are  amplified  and  fed  to  a low  imped- 
ance phase- inverter  to  furnish  balanced  audio. 
This  audio,  together  with  a suitably  chosen 
r-f  signal,  is  mixed  in  a ring  modulator,  made 
up  of  small  germanium  diodes.  Depending 
upon  the  choice  of  frequency  of  the  r-f  oscilla- 
tor, either  the  upper  or  lower  sideband  may  be 
applied  to  the  input  of  the  mechanical  filter. 
The  carrier,  to  some  extent,  has  been  rejected 
by  the  ring  modulator.  Additional  carrier  re- 
jection is  afforded  by  the  excellent  passband 


eoo-3000'ii  zoo-soao'c  «j-«5e/rc  4ss-4stKc.  39ss  kc. 


Figure  17 

BLOCK  DIAGRAM  OF  FILTER  EXCITER  EMPLOYING  A 455-KC. 
MECHANICAL  FILTER  FOR  SIDEBAND  SELECTION 


336  Sideband  Transmission 


THE  RADIO 


l.F.  TRANS. 


rri 

4Xi 

IXL 

Fr-24t  CHANNEL  49  CHYSTAL  « AST.  1 KC. 
FT-241  CHANNEL  50  CRYSTAL  - 462.9KC. 


Figure  18 

SIMPLE  CRYSTAL  LATTICE  FILTER 


FREQUENCY  (K.C.) 


characteristics  of  the  mechanical  filter.  For 
simplicity,  the  mixing  and  filtering  operation 
usually  takes  place  at  a frequency  of  455  kilo- 
cycles. The  single-sideband  signal  appearing 
at  the  output  of  the  mechanical  filter  may  be 
translated  directly  to  a higher  operating  fre- 
quency. Suitable  tuned  circuits  must  follow 
the  conversion  stage  to  eliminate  the  signal 
from  the  conversion  oscillator. 

Wove  Filters  The  heart  of  a filter-type  SSB 
exciter  is  the  sideband  filter. 
Conventional  coils  and  capacitors  may  be 
used  to  construct  a filter  based  upon  standard 
wave  filter  techniques.  The  Q of  the  filter  in- 
ductances must  be  high  when  compared  with 
the  reciprocal  of  the  fractional  bandwidth.  If 
a bandwidth  of  3 kc.  is  needed  at  a carrier  fre- 
quency of  50  kc.,  the  bandwidth  expressed  in 
terms  of  the  carrier  frequency  is  3/50  or  6%. 
This  is  expressed  in  terms  of  fractional  band- 
width as  1/16.  For  satisfactory  operation,  the 


246  247  246  249  260  261  262  266  264  256 


FREQUENCY  (K.C.) 

Figure  1 9 

PASSBAND  OF  LOWER  AND  UPPER 
SIDEBAND  MECHANICAL  FILTER 


Q of  the  filter  inductances  should  be  10  times 
the  reciprocal  of  this,  or  160.  Appropriate  Q is 
generally  obtained  from  toroidal  inductances, 
though  there  is  some  possibility  of  using  iron 
core  solenoids  between  10  kc.  and  20  kc.  A 
characteristic  impedance  below  1000  ohms 
should  be  selected  to  prevent  distributed  ca- 
pacity of  the  inductances  from  spoiling  overall 
performance.  Paper  capacitors  intended  for 
bypass  work  may  not  be  trusted  for  stability 
or  low  loss  and  should  not  be  used  in  filter 
circuits.  Care  should  be  taken  that  the  levels 
of  both  accepted  and  rejected  signals  are  low 
enough  so  that  saturation  of  the  filter  induct- 
ances does  not  occur. 

Crystal  Filters  The  best  known  filter  re- 
sponses have  been  obtained 
with  crystal  filters.  Types  designed  for  pro- 
gram carrier  service  cut-off  80  db  in  less  than 
50  cycles.  More  than  80  crystals  are  used  in 
this  type  of  filter.  The  crystals  are  cut  to  con- 
trol reactance  and  resistance  as  well  as  the 
resonant  frequency.  The  circuits  used  are  based 
on  full  lattices. 

The  war-surplus  low  frequency  crystals  may 
be  adapted  to  this  type  of  filter  with  some 
success.  Experimental  designs  usually  syn- 
thesize a selectivity  curve  by  grouping  sharp 
notches  at  the  side  of  the  passband.  Where  the 
width  of  the  passband  is  greater  than  twice 
the  spacing  of  the  series  and  parallel  reso- 
nance of  the  crystals,  special  circuit  techni- 
ques must  be  used.  A typical  crystal  filter  using 
these  surplus  crystals,  and  its  approximate  pass- 
band  is  shown  in  figure  18. 

Mechanical  Filters  Filters  using  mechanical 
resonators  have  been 
studied  by  a number  of  companies  and  are 
offered  commercially  by  the  Collins  Radio  Co. 
They  ate  available  in  a variety  of  bandwidths 


HANDBOOK 


Generation  of  S.S.B.  337 


TO  POWER  AMPLIFIER  STAGES 
OR  DIRECTLY  TO  ANTENNA  SYSTEM 


Figure  20 

BLOCK  DIAGRAM  OF  THE  "PHASING"  METHOD 

The  phasing  method  of  obtaining  a single-sideband  signal  is  simpler  than  the  filter  system  in  regard  to 
the  number  of  tubes  and  circuits  required.  The  system  is  also  less  expensive  in  regard  to  the  components 
required,  but  is  more  critical  in  regard  to  adjustments  for  the  transmission  of  a pure  single-sideband  signal. 


at  center  frequencies  of  250  kc.  and  455  kc. 
The  250  kc.  series  is  specifically  intended  for 
sideband  selection.  The  selectivity  attained  by 
these  filters  is  intermediate  between  good  LC 
filters  at  low  center  frequencies  and  engineered 
quartz  crystal  filters.  A passband  of  two  250 
kc.  filters  is  shown  in  figure  19.  In  application 
of  the  mechanical  filters  some  special  precau- 
tions are  necessary.  The  driving  and  pick-up 
coils  should  be  carefully  resonated  to  the  op- 
erating frequency.  If  circuit  capacities  are  un- 
known, trimmer  capacitors  should  be  used 
across  the  coils.  Maladjustment  of  these  tuned 
coils  will  increase  insertion  loss  and  the  peak- 
to-valley  ratio.  On  high  impedance  filters  (ten 
to  twenty  thousand  ohms)  signals  greater  than 
2 volts  at  the  input  should  be  avoided.  D-c 
should  be  blocked  out  of  the  end  coils.  While 
the  filters  are  rated  for  5 ma.  of  coil  current, 
they  are  not  rated  for  d-c  plate  voltage. 

The  Phasing  There  are  a number  of  points 
System  of  view  from  which  the  op- 

eration of  the  phasing  system 
of  SSB  generation  may  be  described.  We  may 
state  that  we  generate  two  double-sideband 
suppressed  carrier  signals,  each  in  its  own  bal- 
anced modulator,  that  both  the  r-f  phase  and 
the  audio  phase  of  the  two  signals  differ  by 
90  degrees,  and  that  the  outputs  of  the  two 
balanced  modulators  are  added  with  the  result 
that  one  sideband  is  increased  in  amplitude 
and  the  other  one  is  cancelled.  This,  of  course, 
is  a true  description  of  the  action  that  takes 
place.  But  it  is  much  easier  to  consider  the 
phasing  system  as  a method  simply  of  adding 
(or  of  subtracting)  the  desired  modulation 
frequency  and  the  nominal  carrier  frequency. 
The  carrier  frequency  of  course  is  not  trans- 


mitted, as'  is  the  case  with  all  SSB  transmis- 
sions, but  only  the  sum  or  the  difference  of  the 
modulation  band  from  the  nominal  carrier  is 
transmitted  (figure  20). 

The  phasing  system  has  the  obvious  advan- 
tage that  all  the  electrical  circuits  which  give 
rise  to  the  single  sideband  can  operate  in  a 
practical  transmitter  at  the  nominal  output  fre- 
quency of  the  transmiter.  That  is  to  say  that 
if  we  desire  to  produce  a single  sideband  whose 
nominal  carrier  frequency  is  3-9  Me.,  the 
balanced  modulators  are  fed  with  a 3-9-Mc. 
signal  and  with  the  audio  signal  from  the 
phase  splitters.  It  is  not  necessary  to  go  through 
several  frequency  conversions  in  order  to  ob- 
tain a sideband  at  the  desired  output  fre- 
quency, as  in  the  case  with  the  filter  method 
of  sideband  generation. 

Assuming  that  we  feed  a speech  signal  to 
the  balanced  modulators  along  with  the  3900- 
kc.  carrier  ( 3.9  Me. ) we  will  obtain  in  the  out- 
put of  the  balanced  modulators  a signal  which 
is  either  the  sum  of  the  carrier  signal  and  the 
speech  band,  or  the  difference  between  the  car- 
rier and  the  speech  band.  Thus  if  our  speech 
signal  covers  the  band  from  200  to  3000 
cycles,  we  will  obtain  in  the  output  a band  of 
frequencies  from  3900.2  to  3903  kc.  (the  sum 
of  the  two,  or  the  "upper”  sideband),  or  a band 
from  3897  to  3899.8  kc.  (the  difference  be- 
tween the  two  or  the  "lower”  sideband).  A 
further  advantage  of  the  phasing  system  of 
sideband  generation  is  the  fact  that  it  is  a very 
simple  matter  to  select  either  the  upper  side- 
band or  the  lower  sideband  for  transmission. 
A simple  double-pole  double-throw  reversing 
switch  in  two  of  the  four  audio  leads  to  the 
balanced  modulators  is  all  that  is  required. 


338  Sideband  Transmission 


THE  RADIO 


0«  ia0*«)*270* 

Iv 

FOUR-PHASeA.F. 


Figure  21 

TWO  CIRCUITS  FOR  SINGLE 
SIDEBAND  GENERATION  BY  THE 
PHASING  METHOD. 

The  circuit  of  (A)  offers  the  advantages  of 
simplicity  in  the  single-ended  input  circuits 
plus  a push-pull  output  circuit.  Circuit  (B)  re- 
quires double-ended  input  circuits  but  allows 
all  the  plates  to  be  connected  in  parallel  for 
the  output  circuit. 


High-Level  The  plate-circuit  effi- 

Phosing  Vs.  ciency  of  the  four  tubes 

Low-Level  Phasing  usually  used  to  make  up 
the  two  balanced  modu- 
lators of  the  phasing  system  may  run  as  high 
as  50  to  70  per  cent,  depending  upon  the  oper- 
ating angle  of  plate  current  flow.  Hence  it  is 
possible  to  operate  the  double  balanced  modu- 
lator directly  into  the  antenna  system  as  the 
output  stage  of  the  transmitter. 

The  alternative  arrangement  is  to  generate 
the  SSB  signal  at  a lower  level  and  then  to 
amplify  this  signal  to  the  level  desired  by 
means  of  class  A or  class  B r-f  power  ampli- 
fiers. If  the  SSB  signal  is  generated  at  a level 


of  a few  milliwatts  it  is  most  common  to  make 
the  first  stage  in  the  amplifier  chain  a class 
A amplifier,  then  to  use  one  or  more  class  B 
linear  amplifiers  to  bring  the  output  up  to  the 
desired  level. 

Balanced  Illustrated  in  figure  8 are 

Modulator  Circuits  the  two  basic  balanced 
modulator  circuits  which 
give  good  results  with  a radio  frequency  car- 
rier and  an  audio  modulating  signal.  Note  that 
one  push-pull  and  one  single  ended  tank  cir- 
cuit is  required,  but  that  the  push-pull  circuit 
may  be  placed  either  in  the  plate  or  the  grid 
circuit.  Also,  the  audio  modulating  voltage  al- 
ways is  fed  into  the  stage  in  push-pull,  and 
the  tubes  normally  are  operated  Class  A. 

When  combining  two  balanced  modulators 
to  make  up  a double  balanced  modulator  as 
used  in  the  generation  of  an  SSB  signal  by  the 
phasing  system,  only  one  plate  circuit  is  re- 
quired for  the  two  balanced  modulators.  How- 
ever, separate  grid  circuits  are  required  since 
the  grid  circuits  of  the  two  balanced  modula- 
tors operate  at  an  r-f  phase  difference  of  90 
degrees.  Shown  in  figure  21  are  the  two  types 
of  double  balanced  modulator  circuits  used  for 
generation  of  an  SSB  signal.  Note  that  the  cir- 
cuit of  figure  21 A is  derived  from  the  bal- 
anced modulator  of  figure  8A,  and  similarly 
figure  2 IB  is  derived  from  figure  8B. 

Another  circuit  that  gives  excellent  perform- 
ance and  is  very  easy  to  adjust  is  shown  in 
figure  22.  The  adjustments  for  carrier  balance 
are  made  by  adjusting  the  potentiometer  for 
voltage  balance  and  then  the  small  variable  ca- 
pacitor for  exact  phase  balance  of  the  balanced 
carrier  voltage  feeding  the  diode  modulator. 


Figure  22 

BALANCED  MODULATOR  FOR  USE 
WITH  MECHANICAL  FILTER 


HANDBOOK 


Generation  of  S.S.B.  339 


is  convenient  in  a case  where  it  is  desired  to 
make  small  changes  in  the  operating  /re* 
quency  of  the  system  without  the  necessity 
of  being  precise  in  the  adjustment  of  two 
coupled  circuits  as  used  for  r-f  phase  shift 
in  the  circuit  of  figure  21. 


Radio-Frequency  A single-sideband  genera- 

Phasing  tor  of  the  phasing  type 

requites  that  the  two  bal- 
anced modulators  be  fed  with  r-f  signals  hav- 
ing a 90-degtee  phase  difference.  This  r-f 
phase  difference  may  be  obtained  through  the 
use  of  two  loosely  coupled  resonant  circuits, 
such  as  illustrated  in  figures  21A  and  21B. 
The  r-f  signal  is  coupled  directly  or  inductive- 
ly to  one  of  the  tuned  circuits,  and  the  coupling 
between  the  two  circuits  is  varied  until,  at 
resonance  of  both  circuits,  the  r-f  voltages 
developed  across  each  circuit  have  the  same 
amplitude  and  a 90-degree  phase  difference. 

The  90-degree  r-f  phase  difference  also  may 
be  obtained  through  the  use  of  a low-Q  phase 
shifting  network,  such  as  illustrated  in  figure 
23;  or  it  may  be  obtained  through  the  use  of 
a lumped-constant  quarter-wave  line.  The  low- 
Q phase-shifting  system  has  proved  quite  prac- 
ticable for  use  in  single-sideband  systems, 
particularly  on  the  lower  frequencies.  In  such 
an  arrangement  the  two  resistances  R have 
the  same  value,  usually  in  the  range  between 
100  and  a few  thousand  ohms.  Capacitor  C,  in 
shunt  with  the  input  capacitances  of  the  tubes 
and  circuit  capacitances,  has  a reactance  at 
the  operating  frequency  equal  to  the  value  of 
the  resistor  R.  Also,  inductor  L has  a net  in- 
ductive reactance  equal  in  value  at  the  oper- 
ating frequency  to  resistance  R. 

The  inductance  chosen  for  use  at  L must 
take  into  account  the  cancelling  effect  of  the 
input  capacitance  of  the  tubes  and  the  circuit 
capacitance;  hence  the  inductance  should  be 


taining  the  audio  phase  shift  when  it  is  desired 
to  use  a minimum  of  circuit  components  and 
tube  elements. 


variable  and  should  have  a lower  value  of  in- 
ductance  than  that  value  of  inductance  which 
would  have  the  same  reactance  as  resistor  R. 
Inductor  L may  be  considered  as  being  made 
up  of  two  values  of  inductance  in  parallel;  (a) 
a value  of  inductance  which  will  resonate  at 
the  operating  frequency  with  the  circuit  and 
tube  capacitances,  and  ( b ) the  value  of  induct- 
ance which  is  equal  in  reactance  to  the  resist- 
ance R.  In  a network  such  as  shown  in  figure 
23,  equal  and  opposite  45-degtee  phase  shifts 
are  provided  by  the  RL  and  RC  circuits,  thus 
providing  a 90-degree  phase  difference  be- 
tween the  excitation  voltages  applied  to  the 
two  balanced  modulators. 

Audio-Frequency  The  audio-frequency  phase- 

Phasing  shifting  networks  used  in 

generating  a single-side- 
band signal  by  the  phasing  method  usually  are 
based  on  those  described  by  Dome  in  an  ar- 
ticle in  the  December,  1946,  Electronics.  A 
relatively  simple  network  for  accomplishing 
the  90-degree  phase  shift  over  the  range  from 
160  to  3500  cycles  is  illustrated  in  figure  24. 
The  values  of  resistance  and  capacitance  must 
be  carefully  checked  to  insure  minimum  devia- 
tion from  a 9b-degree  phase  shift  over  the  200 
to  3000  cycle  range. 

Another  version  of  the  Dome  network  is 
shown  in  figure  25.  This  network  employs 
three  12AU7  tubes  and  provides  balanced  out- 
put for  the  two  balanced  modulators.  As  with 
the  previous  network,  values  of  the  resistances 
within  the  network  must  be  held  to  very  close 
tolerances.  It  is  necessary  to  restrict  the  speech 

range  to  300  to  3000  cycles  with  this  network. 
Audio  frequencies  outside  this  range  will  not 
have  the  necessary  phase-shift  at  the  output 


340  Sideband  Transmission 


THE  RADIO 


12AU7  I2AU7  12AU7 


AUDIO-PHASE-SHIFT 

NETWORK 


of  the  network  and  will  show  up  as  spurious 
emissions  on  the  sideband  signal,  and  also 
in  the  region  of  the  rejected  sideband.  A low- 
pass  3500  cycle  speech  filter,  such  as  the 
Chicago  Transformer  Co.  LPF-2  should  be 
used  ahead  of  this  phase-shift  network. 

A passive  audio  phase-shift  network  that 
employs  no  tubes  is  shown  in  figure  26.  This 
network  has  the  same  type  of  operating  re- 
strictions as  those  described  above.  Additional 
information  concerning  phase-shift  networks 
will  be  found  in  Single  Sideband  T echniques 
published  by  rhe  Cowan  Publishing  Corp., 
New  York,  and  The  Single  Sideband  Digest 
published  by  the  American  Radio  Relay 
League.  A comprehensive  sideband  review  is 
contained  in  the  December,  1956  issue  of 
Proceedings  of  the  I.R.E- 

Comparison  of  Filter  Either  the  filter  or  the 
and  Phasing  Methods  phasing  method  of 

of  SSB  Generation  single-sideband  gener- 

ation is  theoretically 
capable  of  a high  degree  of  performance. 

In  general,  it  may  be  said  that  a high  degree 
of  unwanted  signal  rejection  may  be  attained 
with  less  expense  and  circuit  complexity  with 
the  filter  method.  The  selective  circuits  for 
rejection  of  unwanted  frequencies  operate  at  a 
relativly  low  frequency,  are  designed  for  this 
one  frequency  and  have  a relatively  high  order 
of  Q.  Carrier  rejection  of  the  order  of  50  db  or 
so  may  be  obtained  with  a relatively  simple 
filter  and  a balanced  modulator,  and  unwanted 
sideband  rejection  in  the  region  of  60  db  is 
economically  possible. 

The  phasing  method  of  SSB  generation  ex- 
changes the  problems  of  high-Q  circuits  and 
linear  amplification  for  the  problems  of  accu- 
rately controlled  phase-shift  networks.  If  the 


PASSIVE  AUDIO-PHASE-SHIFT 
NETWORK,  USEFUL  OVER  RANGE 
OF  300  TO  3000  CYCLES. 


phasing  method  is  employed  on  the  actual 
transmitting  frequency,  change  of  frequency 
must  be  accompanied  by  a corresponding  re- 
balance of  the  phasing  networks.  In  addition, 
it  is  difficult  to  obtain  a phase  balance  with 
ordinary  equipment  within  1%  over  a band  of 
audio  frequencies.  This  means  that  carrier 
suppression  is  limited  to  a maximum  of  40  db 
or  so.  However,  when  a relatively  simple  SSB 
transmitter  is  needed  for  spot  frequency  opera- 
tion, a phasing  unit  will  perform  in  a satis- 
factory manner. 

Where  a high  degree  of  performance  in  the 
SSB  exciter  is  desired,  the  filter  method 
and  the  phasing  method  may  be  combined. 
Through  the  use  of  the  phasing  method  in  the 
first  balanced  modulator  those  undesired  side- 
band components  lying  within  1000  cycles  of 
the  carrier  may  be  given  a much  higher  degree 
of  rejection  than  is  attainable  with  the  filter 
method  alone,  with  any  reasonable  amount  of 
complexity  in  the  sideband  filter.  Then  the 
sideband  filter  may  be  used  in  its  normal  way 
to  attain  very  high  attenuation  of  all  undesired 
sideband  components  lying  perhaps  further 
than  500  cycles  away  from  the  carrier,  and  to 
restrict  the  sideband  width  on  the  desired  side 
of  the  carrier  to  the  specified  frequency  limit. 

17-5  Single  Sideband 
Frequency  Conversion  Systems 

In  many  instances  the  band  of  sideband 
frequencies  generated  by  a low  level  SSB  trans- 
mitter must  be  heterodyned  up  to  the  desired 
carrier  frequency.  In  receivers  the  circuits 
which  perform  this  function  are  called  con- 
verters or  mixers.  In  sideband  work  they  are 
usually  termed  mixers  or  modulators. 

Mixer  Stages  One  circuit  which  can  be  used 
for  this  purpose  employs  a 
receiving-type  mixer  tube,  such  as  the  6BE6. 
The  output  signal  from  the  SSB  generator  is 
fed  into  the  #1  grid  and  the  conversion  fre- 


HANDBOOK 


Frequency  Conversion  341 


6BE6 


- hL-ILJ 

n 

o 

o 

'•np 

250  KC.  SSB 
SIGNAL 
(0.25  K ) 


.TUNE  TO  SELECT 
2000  + 250=2250  KC 


-2000-  250=1750  KC. 


Figure  27 

PENTAGRID  MIXER  CIRCUIT  FOR 
SSB  FREQUENCY  CONVERSION 


quency  into  the  # 3 grid.  This  is  the  reverse  of 
the  usual  grid  connections,  but  it  offers  about 
10  db  improvement  in  distortion.  The  plate 
circuit  is  tuned  to  select  the  desired  output 
frequency  product.  Actually,  the  output  of  the 
mixer  tube  contains  all  harmonics  of  the  two 
input  signals  and  all  possible  combinations  of 
the  sum  and  difference  frequencies  of  all  the 
harmonics.  In  order  to  avoid  distortion  of  the 
SSB  signal,  it  is  fed  to  the  mixer  at  a low 
level,  such  as  0.1  to  0.2  volts.  The  conversion 
frequency  is  fed  in  at  a level  about  20  db 
higher,  or  about  2 volts.  By  this  means,  har- 
monics of  the  incoming  SSB  signal  generated 
in  the  mixer  tube  will  be  very  low.  Usually 
the  desired  output  frequency  is  either  the  sum 
or  the  difference  of  the  SSB  generator  carrier 
frequency  and  the  conversion  frequency.  For 
example,  using  a SSB  generator  carrier  fre- 
quency of  250  kc.  and  a conversion  injeaion 
frequency  of  2000  kc.  as  shown  in  figure  27, 
the  output  may  be  tuned  to  select  either  2250 
kc.  or  1750  kc. 

Not  only  is  it  necessary  to  select  the  desired 
mixing  product  in  the  mixer  output  but  also 
the  undesired  products  must  be  highly  attenu- 
ated to  avoid  having  spurious  output  signals 
from  the  transmitter.  In  general,  all  spurious 
signals  that  appear  within  the  assigned  fre- 
quency channel  should  be  at  least  60  db  below 
the  desired  signal,  and  those  appearing  out- 
side of  the  assigned  frequency  channel  at  least 
80  db  below  the  signal  level. 

When  mixing  250  kc.  with  2000  kc.  as  in 
the  above  example,  the  desired  product  is  the 
2250  kc.  signal,  but  the  2000  kc.  injection 
frequency  will  appear  in  the  output  about  20 
db  stronger  than  the  desired  signal.  To  reduce 
it  to  a level  80  db  below  the  desired  signal 
means  that  it  must  be  attenuated  100  db. 

The  principal  advantage  of  using  balanced 
modulator  mixer  stages  is  that  the  injection 
frequency  theoretically  does  not  appear  in  the 
output.  In  practice,  when  a considerable  fre- 
quency range  must  be  tuned  by  the  balanced 
modulator  and  it  is  not  practical  to  trim  the 


TWIN  TRIODE  MIXER  CIRCUIT  FOR 
SSB  FREQUENCY  CONVERSION 


push-pull  circuits  and  the  tubes  into  exact 
amplitude  and  phase  balance,  about  20  db 
of  injection  frequency  cancellation  is  all  that 
can  be  depended  upon.  With  suitable  trim- 
ming adjustments  the  cancellation  can  be  made 
as  high  as  40  db,  however,  in  fixed  frequency 
circuiis. 

The  Twin  Triode  Mixer  The  mixer  circuit 
shown  in  figure  28 
has  about  10  db  lower  distortion  than  the  con- 
ventional 6BE6  converter  tube.  It  has  a lower 
voltage  gain  of  about  unity  and  a lower  out- 
put impedance  which  loads  the  first  tuned 
circuit  and  reduces  its  selectivity.  In  some  ap- 
plications the  lower  gain  is  of  no  consequence 
but  the  lower  distortion  level  is  important 
enough  to  warrant  its  use  in  high  performance 
equipment.  The  signal-to-distortion  ratio  of 
this  mixer  is  of  the  order  of  70  db  compared 
to  approximately  60  db  for  a 6BE6  mixer 
when  the  level  of  each  of  two  tone  signals  is 
0.5  volt.  With  stronger  signals,  the  6BE6 
distortion  increases  very  rapidly,  whereas  the 
12AU7  distortion  is  much  better  compara- 
tively. 


6AS6's 


Figure  29 

BALANCED  MODULATOR  CIRCUIT 
FOR  SSB  FREQUENCY  CONVERSION 


f CYCLES  OFF  RESONANCE 
77“  RESONANT  FREQUENCY 


HANDBOOK 


Frequency  Conversion  343 


In  practical  equipment  where  the  injection 
frequency  is  variable  and  trimming  adjust- 
ments and  tube  selection  cannot  be  used,  it 
may  be  easier  and  more  economical  to  obtain 
this  extra  20  db  of  attenuation  by  using  an 
extra  tuned  circuit  in  the  output  than  by  using 
a balanced  modulator  circuit.  A balanced  mod- 
ulator circuit  of  interest  is  shown  in  figure 
29,  providing  a minimum  of  20  db  of  carrier 
attenuation  with  no  balancing  adjustment. 

Selective  Tuned  Circuits  The  selectivity  re- 
quirements of  the 
tuned  circuits  following  a mixer  stage  often 
become  quite  severe.  For  example,  using  an 
input  signal  at  250  kc.  and  a conversion  in- 
jection frequency  of  4000  kc.  the  desired  out- 
put may  be  4250  kc.  Passing  the  4250  kc. 
signal  and  the  associated  sidebands  without 
attenuation  and  realizing  100  db  of  attenuation 
at  4000  kc.  (which  is  only  250  kc.  away)  is 
a practical  example.  Adding  the  requirement 
that  this  selective  circuit  must  tune  from  2250 
kc.  to  4250  kc.  further  complicates  the  basic 
requirement.  The  best  solution  is  to  cascade  a 
number  of  tuned  circuits.  Since  a large  num- 
ber of  such  circuits  may  be  required,  the  most 
practical  solution  is  to  use  permeability  tun- 
ing, with  the  circuits  tracked  together.  An  ex- 
ample of  such  circuitry  is  found  in  the  Collins 
KWS-1  sideband  transmitter. 

If  an  amplifier  rube  is  placed  between  each 
tuned  circuit,  the  overall  response  will  be  the 
sum  of  one  stage  multiplied  by  the  number 
of  stages  (assuming  identical  tuned  circuits). 
Figure  30  is  a chart  which  may  be  used  to 
determine  the  number  of  tuned  circuits  re- 
quired for  a certain  degree  of  attenuation  at 
some  nearby  frequency.  The  Q of  the  circuits 
is  assumed  to  be  50,  which  is  normally  realized 
in  small  permeability  tuned  coils.  The  number 
of  tuned  circuits  with  a Q of  50  required  for 
providing  100  db  of  attenuation  at  4000  kc. 
while  passing  4250  kc.  may  be  found  as  fol- 
lows: 


Af  is  4250-4000=250  kc. 

fr  is  the  resonant  frequency,  4250  kc. 


250 

4250 


0.059 


The  point  on  the  chart  where  .059  inter- 
sects 100  db  is  between  the  curves  for  6 and  7 
tuned  circuits,  so  7 tuned  circuits  ate  required. 

Another  point  which  must  be  considered  in 
practice  is  the  mning  and  tracking  error  of 
the  circuits.  For  example,  if  the  circuits  were 


actually  tuned  to  4220  kc.  instead  of  4250  kc., 
Af  220 

the  ^ — would  be  ^220  0.0522.  Checking 

the  curves  shows  that  7 circuits  would  just 
barely  provide  100  db  of  attenuation.  This 
illustrates  the  need  for  very  accurate  tuning 
and  tracking  in  circuits  having  high  attenua- 
tion properties. 

Coupled  Tuned  When  as  many  as  7 tuned 

Circuits  circuits  are  required  for  pro- 

per attenuation,  it  is  not 
necessary  to  have  the  gain  that  6 isolating  am- 
plifier tubes  would  provide.  Several  vacuum 
tubes  can  be  eliminated  by  using  two  or  three 
coupled  circuits  between  the  amplifiers.  With 
a coefficient  of  coupling  between  circuits  0.5 
of  critical  coupling,  the  overall  response  is 
very  nearly  the  same  as  isolated  circuits.  The 
gain  through  a pair  of  circuits  having  0-5 
coupling  is  only  eight-tenths  that  of  two  criti- 
cally coupled  circuits,  however.  If  critical 
coupling  is  used  between  two  tuned  circuits, 
the  nose  of  the  response  curve  is  broadened 
and  about  6 db  is  lost  on  the  skirts  of  each 
pair  of  critically  coupled  circuits.  In  some 
cases  it  may  be  necessary  to  broaden  the  nose 
of  the  response  curve  to  avoid  adversely  af- 
fecting the  frequency  response  of  the  desired 
passband.  Another  tuned  circuit  may  be  re- 
quired to  make  up  for  the  loss  of  attenuation 
on  the  skirts  of  critically  coupled  circuits. 

Frequency  Conversion  The  example  in  the 

Problems  previous  section  shows 

the  difficult  selectivi- 
ty problem  encountered  when  strong  undesired 
signals  appear  near  the  desired  frequency.  A 
high  frequency  SSB  transmitter  may  be  re- 
quired to  operate  at  any  carrier  frequency  in 
the  range  of  1.75  Me.  to  30  Me.  The  problem 
is  to  find  a praaical  and  economical  means  of 
heterodyning  the  generated  SSB  frequency  to 
any  carrier  frequency  in  this  range.  There  are 
many  modulation  products  in  the  output  of  the 
mixer  and  a frequency  scheme  must  be  found 
that  will  not  have  undesired  output  of  appre- 
ciable amplitude  at  or  near  the  desired  signal. 
When  tuning  across  a frequency  range  some 
products  may  "cross  over”  the  desired  fre- 
quency. These  undesired  crossover  frequencies 
should  be  at  least  60  db  below  the  desired 
signal  to  meet  modern  standards.  The  ampli- 
tude of  the  undesired  products  depends  upon 
the  particular  characteristics  of  the  mixer  and 

the  particular  order  of  the  product.  In  general, 

most  products  of  the  7th  order  and  higher 
will  be  at  least  60  db  down.  Thus  any  cross- 


344  Sideband  Transmission 


THE  RADIO 


FROM  SSB 
GENERATOR 


DELAY  BIAS  VOLTAGE 
FROM  POWER  SUPPLY 


Figure  31  Figure  32 

SSB  DISTORTION  PRODUCTS,  BLOCK  DIAGRAM  OF  AUTOMATIC 

SHOWN  UP  TO  NINTH  ORDER  LOAD  CONTROL  (A.L.C.)  SYSTEM 


over  frequency  lower  than  the  7th  must  be 
avoided  since  there  is  no  way  of  attenuating 
them  if  they  appear  within  the  desired  pass- 
band.  The  General  Electric  Ham  News,  volume 
11  #6  of  Nov.-Dee.,  1956  covers  the  subject 
of  spurious  products  and  incorporates  a "mix- 
selector”  chart  that  is  useful  in  determining 
spurious  products  for  various  different  mixing 
schemes. 

In  general,  for  most  applications  when  the 
intelligence  bearing  frequency  is  lower  than 
the  conversion  frequency,  it  is  desirable  that 
the  ratio  of  the  two  frequencies  be  between  5 
to  1 and  10  to  1.  This  is  a compromise  be- 
tween avoiding  low  order  harmonics  of  this 
signal  input  appearing  in  the  output,  and 
minimizing  the  selectivity  requirements  of  the 
circuits  following  the  mixer  stage. 

17-6  Distortion  Products 
Due  to  Nonlinearity  of 
R-F  Amplifiers 

When  the  SSB  envelope  of  a voice  signal 
is  distorted,  a great  many  new  frequencies  are 
generated.  These  represent  all  of  the  possible 
combinations  of  the  sum  and  difference  fre- 
quencies of  all  harmonics  of  the  original  fre- 
quencies. For  purposes  of  test  and  analysis, 
two  equal  amplimde  tones  are  used  as  the 
SSB  audio  source.  Since  the  SSB  radio  fre- 
quency amplifiers  use  tank  circuits,  all  distor- 
tion products  ate  filtered  out  except  those 
which  lie  close  to  the  desired  frequencies. 
These  are  all  odd  order  products;  third  order, 
fifth  order,  etc..  The  third  order  products  are 
2p-q  and  2q-p  where  p and  q represent  the 
two  SSB  r-f  tone  frequencies.  The  fifth  order 
products  are  3p-2q  and  3q-2p.  These  and 
some  higher  order  products  are  shown  in 
figure  31-  It  should  be  noted  that  the  fre- 
quency spacings  are  always  equal  to  the  dif- 
ference frequency  of  the  two  original  tones. 
Thus  when  a SSB  amplifier  is  badly  over- 


loaded, these  spurious  frequencies  can  extend 
far  outside  the  original  channel  width  and 
cause  an  unintelligible  "splatter”  type  of  in- 
terference in  adjacent  channels.  This  is  usually 
of  far  more  importance  than  the  distortion  of 
the  original  tones  with  regard  to  intelligibility 
or  fidelity.  To  avoid  interference  in  another 
channel,  these  distortion  products  should  be 
down  at  least  40  db  below  adjacent  channel 
signal.  Using  a two-tone  test,  the  distortion  is 
given  as  the  ratio  of  the  amplitude  of  one 
test  tone  to  the  amplitude  of  a third  order 
product.  This  is  called  the  signal-to-distortion 
ratio  (S/D)  and  is  usually  given  in  decibels. 
The  use  of  feedback  r-f  amplifiers  make  S/D 
ratios  of  greater  than  40  db  possible  and  prac- 
tical. 

Automatic  Two  means  may  be  used  to 

Load  Control  keep  the  amplitude  of  these 
distortion  products  down  to 
acceptable  levels.  One  is  to  design  the  ampli- 
fier for  excellent  linearity  over  its  amplitude 
or  power  range.  The  other  is  to  employ  a 
means  of  limiting  the  amplitude  of  the  SSB 
envelope  to  the  capabilities  of  the  amplifier. 
An  automatic  load  control  ssytem  (ALC)  may 
be  used  to  accomplish  this  result.  It  should  be 
noted  that  the  r»f  wave  shapes  of  the  SSB  sig- 
nal are  always  sine  waves  because  the  tank  cir- 
cuits make  them  so.  It  is  the  change  in  gain 
with  signal  level  in  an  amplifier  that  distorts 
the  SSB  envelope  and  generates  unwanted  dis- 
tortion products.  An  ALC  system  may  be  used 
to  limit  the  input  signal  to  an  amplifier  to 
prevent  a change  in  gain  level  caused  by  ex- 
cessive input  level. 

The  ALC  system  is  adjusted  so  the  power 
amplifier  is  operating  near  its  maximum  power 
capability  and  at  the  same  time  is  protected 
from  being  over-driven.  In  amplitude  modu- 
lated systems  it  is  common  to  use  speech  com- 
pressors and  speech  clipping  systems  to  per- 
form this  function.  These  methods  ate  not 


HANDBOOK 


Distortion  Products  345 


Figure  33 

PERFORMANCE  CURVE  OF 
A.LC.  CIRCUIT 


equally  useful  in  SSB.  The  reason  for  this  is 
that  the  SSB  envelope  is  different  from  the 
audio  envelope  and  the  SSB  peaks  do  not 
necessarily  correspond  with  the  audio  peaks 
as  explained  earlier  in  this  chapter.  For  this 
reason  a "compressor”  of  some  sort  located 
between  the  SSB  generator  and  the  power  am- 
plifier is  most  effective  because  it  is  controlled 
by  SSB  envelope  peaks  rather  than  audio  peaks. 
Such  a "SSB  signal  compressor”  and  the  means 
of  obtaining  its  control  voltage  comprises  a 
satisfactory  ALC  system. 

The  ALC  Circuit  A block  diagram  of  an  ALC 
circuit  is  shown  in  figure 
32.  The  compressor  or  gain  control  part  of 
this  circuit  uses  one  or  two  stages  of  remote 
cutoff  tubes  such  as  6BA6,  operating  very  sim- 
ilarly to  the  intermediate  frequency  stages  of 
a receiver  having  automatic  volume  control. 

The  grid  bias  voltage  which  controls  the 
gain  of  the  tubes  is  obtained  from  a voltage 
detector  circuit  connected  to  the  power  am- 
plifier tube  plate  circuit.  A large  delay  bias  is 
used  so  that  no  gain  reduction  takes  place  until 


the  signal  is  nearly  up  to  the  full  power  capa- 
bility of  the  amplifier.  At  this  signal  level, 
the  rectified  output  overcomes  the  delay  bias 
and  the  gain  of  the  preamplifier  is  reduced 
rapidly  with  increasing  signal  so  that  there  is 
very  little  rise  in  output  power  above  the 
threshold  of  gain  control. 

When  a signal  peak  arrives  that  would  nor- 
mally overload  the  power  amplifier,  it  is  de- 
sireable  that  the  gain  of  the  ALC  amplifier  be 
reduced  in  a few  milliseconds  to  a value  where 
overloading  of  the  power  amplifier  is  over- 
come. After  the  signal  peak  passes,  the  gain 
should  return  to  the  normal  value  in  about 
one-tenth  second.  These  attack  and  release 
times  are  commonly  used  for  voice  communi- 
cations. For  this  type  of  work,  a dynamic  range 
of  at  least  10  db  is  desirable.  Input  peaks  as 
high  as  20  db  above  the  threshold  of  com- 
pression should  not  cause  loss  of  control  al- 
though some  increase  in  distortion  in  the  up- 
per range  of  compression  can  be  tolerated  be- 
cause peaks  in  this  range  are  infrequent.  An- 
other limitation  is  that  the  preceding  SSB 
generator  must  be  capable  of  passing  signals 
above  full  power  output  by  the  amount  of 
compression  desired.  Since  the  signal  level 
through  the  SSB  generator  should  be  main- 
tained within  a limited  range,  it  is  unlikely 
that  more  than  12  db  ALC  action  will  be 
useful.  If  the  input  signal  varies  more  than 
this,  a speech  compressor  should  be  used  to 
limit  the  range  of  the  signal  fed  into  the  SSB 
generator. 

Figure  33  shows  the  effectiveness  of  the 
ALC  in  limiting  the  output  signal  to  the  cap- 
abilities of  the  power  amplifier.  An  adjustment 
of  the  delay  bias  will  place  the  threshold  of 
compression  at  the  desired  power  output.  Fig- 
ure 34  shows  a simplified  schematic  of  an  ALC 
system.  This  ALC  uses  two  variable  gain  am- 


Figure  34 

SIMPLIFIED  SCHEMA- 
TIC OF  AUTOMATIC 
LOAD  CONTROL  AM- 
PLIFIER. OPERATING 
POINT  OF  ALC 
CIRCUIT  MAY  BE 
SET  BY  VARYING 
BLOCKING  BIAS  ON 
CATHODE  OF  6X4 
SIGNAL  RECTIFIER 


346  Sideband  Transmission 


THE  RADIO 


Figure  35 

SSB  JR.  MODULATOR  CIRCUIT 

R-F  and  A-F  sources  are  applied  in  series 
to  balanced  modulator. 


plifier  stages  and  the  maximum  overall  gain  is 
about  20  db.  A meter  is  incorporated  which  is 
calibrated  in  db  of  compression.  This  is  use- 
ful in  adjusting  the  gain  for  the  desired 
amount  of  load  control.  A capacity  voltage 
divider  is  used  to  step  down  the  r-f  voltage 
at  the  plate  of  the  amplifier  tube  to  about  50 
volts  for  the  ALC  rectifier.  The  output  of  the 
ALC  rectifier  passes  through  R-C  networks 
to  obtain  the  desired  attack  and  release  times 


and  through  r-f  filter  capacitors.  The  3.3K 
resistor  and  0.1  gfd.  capacitor  across  the  recti- 
fier output  stabilizes  the  gain  around  the  ALC 
loop  to  prevent  '‘motor-boating.” 

17-7  Sideband  Exciters 

Some  of  the  most  popular  sideband  exciters 
in  use  today  are  variations  of  the  simple  phas- 
ing circuit  introduced  in  the  November,  1950 
issue  of  General  Electric  Ham  News.  Called  the 
SSB,  Jr.,^  this  simple  exciter  is  the  basis  for 
many  of  the  phasing  transmitters  now  in  use. 
Employing  only  three  tubes,  the  SSB,  Jr.  is  a 
classic  example  of  sideband  generation  re- 
duced to  its  simplest  form. 

The  SSB,  Jr.  This  phasing  exciter  employs 
audio  and  r-f  phasing  circuits 
to  produce  a SSB  signal  at  one  spot  frequency. 
The  circuit  of  one  of  the  balanced  modulator 
stages  is  shown  in  figure  35.  The  audio  signal 
and  r-f  source  are  applied  in  series  to  two  ger- 
manium diodes  serving  as  balanced  modulators 


C2A,B.C,D  = EACH  SECTION  ZO  JJF,  450  V. 
C7  = 2430  JLJJLIFD  (.002JUFD  MICA  ±5 9b  WITH 
06=4660  JJiJFD.  (.0043UFDMICA  WITH 

C9=  1215  iJiJFD.(.001  jUFO  MICA  ±5  96  WITH 
C10=607.5JJJUFD  (SOOUUFO  MICA  ±596  WITH 
Cl8=350JLUJF0  600  V.  MICA  ±10  96  (250JLIJUFD 
R7,Ri0=  133.300  OHMS,  1/2  WATT  ± 1 9b 
R8,R9=  100,000  ohms,  1/2  watt  ±196 
Tl  =Sr/iNCOP  A-53C  TRANSFORMER  . 
Tzjz-urc  R-3dA  TRANSFORMER. 

Si  = OPDT  TOGGLE  SWITCH 


ELECTROLYTIC  Gl,2,3,4=  1N52  GERMANIUM  DIODE  OR  EQUIVALENT 

170-760  JUUFD  TRIMMER)  Li,  L2=  33  T.  N»  21  E.  WIRE  CLOSE  WOUND  ON  MILLEN  N»  69046 
170-760  UJUFD  TRIMMER)  IRON  CORE  ADJUSTABLE  SLUG  COI L FORM.  LINK0F6 

50-380 UUFD TRIMMER)  TURNS  OF  HOOKUP  WIRE  WOUND  ON  OPEN  END. 

9-160  JJUFDTRIMMEB)L3=ier.  N«19  E.  WIRE  SPACED  TO  FILL  MILLEN  69046 
AND  100  JUJJFD  PARALLEL)  COIL  FORM.  TAP  AT  8 TURNS.  LINK  OF  1 TURN  AT  CENTER. 

L4=SAME  AS  L1  EXCEPT  NO  LINK  USED. 

L5=  28  T.  OF  N*19  E.  WIRE.  LINK  ON  END  TO  MATCH  LOAD. 

(4  TURN  LINK  MATCHES  72  OHM  LOAD) 

■H-  = MOUNTING  END  OF  COILS. 


Figure  36 
SCHEMATIC,  SSB,  JR. 


HANDBOOK 


S.B.  Exciters  347 


having  a push-pull  output  circuit  tuned  to  the 
r-f  "carrier”  frequency.  The  modulator  drives 
a linear  amplifier  directly  at  the  output  fre- 
quency. The  complete  circuit  of  the  exciter 
is  shown  in  figure  36. 

The  first  tube,  a 12AU7,  is  a twin-triode 
serving  as  a speech  amplifier  and  a crystal 
oscillator.  The  second  tube  is  a 12AT7,  act- 
ing as  a twin  channel  audio  amplifier  follow- 
ing the  phase-shift  audio  network.  The  linear 
amplifier  stage  is  a 6AG7,  capable  of  a peak 
power  output  of  5 watts. 

Sideband  switching  is  accomplished  by  the 
reversal  of  audio  polarity  in  one  of  the  audio 
channels  (switch  Si),  and  provision  is  made 
for  equalization  of  gain  in  the  audio  channels 
(R12).  This  adjustment  is  necessary  in  order 
to  achieve  normal  sideband  cancellation,  which 
may  be  of  the  order  of  35  db  or  better.  Phase- 
shift  network  adjustment  may  be  achieved  by 
adjusting  potentiometer  Rs.  Stable  modulator 
balance  is  achieved  by  the  balance  potentio- 
meters Ri6  and  Ri7  in  conjunction  with  the 
germanium  diodes. 

The  SSB,  Jr.  is  designed  for  spot  frequency 
operation.  Note  that  when  changing  frequency 
Li,  L2,  La,  Li,  and  L«  should  be  readjusted, 
since  these  circuits  constitute  the  tuning  ad- 
justments of  the  tig.  The  principal  effect  of 
mistuning  Li,  L«,  and  L»  will  be  lower  output. 
The  principal  effect  of  mistuning  Lj,  however, 
will  be  degraded  sideband  suppression. 

Power  requirements  of  the  SSB,  Jr.  are  300 
^olts  at  60  ma.,  and  10.5  volts  at  1 ma. 
Under  load  the  total  plate  current  will  rise  to 
about  80  ma.  at  full  level  with  a single  tone 
input.  With  speech  input,  the  total  current 
will  rise  from  the  resting  value  of  60  ma.  to 
about  70  ma.,  depending  upon  the  voice  wave- 
form. 

The  "Ten-A"  The  Model  10- A phasing  ex- 
Exciter  citer  produced  by  Cenpral 

Electronics,  Inc.  is  an  ad- 
vanced version  of  the  SSB,  Jr.  incorporating 
extra  features  such  as  VFO  control,  voice  op- 
eration, and  multi-band  operation.  A simpli- 
fied schematic  of  the  Model  10 A is  shown  in 
figure  37.  The  12AX7  two  stage  speech  am- 
plifier excites  a transformer  coupled  14-12BH7 
low  impedance  driver  stage  and  a voice  oper- 
ated (VOX)  relay  system  employing  a 12AX7 
and  a 6AL5.  A transformer  coupled  12AT7 
follows  the  audio  phasing  network,  providing 
two  audio  channels  having  a 90-clegree  phase 
difference.  A simple  90-degree  r-f  phase  shift 
network  in  the  plate  circuit  of  the  9 Me.  crys- 


tal oscillator  stage  works  into  the  matched, 
balanced  modulator  consisting  of  four  1N48 
diodes. 

The  resulting  9 Me.  SSB  signal  may  be  con- 
verted to  the  desired  operating  frequency  in  a 
6BA7  mixer  stage.  Eight  volts  of  r-f  from  an 
external  v-f-o  injected  on  grid  #1  of  the 
6BA7  is  sufficient  for  good  conversion  effic- 
iency and  low  distorrion.  The  plate  circuit  of 
the  6BA7  is  tuned  to  the  sum  or  difference 
mixing  frequency  and  the  resulting  signal  is 
amplified  in  a 6AG7  linear  amplifier  stage. 
Two  "tweet”  traps  are  incorporated  in  the 
6BA7  stage  to  reduce  unwanted  responses  of 
rhe  mixer  which  are  apparent  when  the  unit  is 
operating  in  the  14  Me.  band.  Band-changing 
is  accomplished  by  changing  coils  Ls  and  Lg 
and  the  frequency  of  the  external  mixing  sig- 
nal. Maximum  power  output  is  of  the  order 
of  5 watts  at  any  operating  frequency. 

A Simple  80  Meter  A SSB  exciter  employ- 
Phasing  Exciter  ing  r-f  and  audio  phas- 

ing circuits  is  shown 
in  figure  38.  Since  the  r-f  phasing  circuits  ate 
balanced  only  at  one  frequency  of  operation, 
the  phasing  exciter  is  necessarily  a single  fre- 
quency transmitter  unless  provisions  ate  made 
to  re-balance  the  phasing  circuits  every  time  a 
frequency  shift  is  made.  However  for  mobile 
operation,  or  spot  frequency  operation  a rela- 
tively simple  phasing  exciter  may  be  made  to 
perform  in  a satisfactory  manner. 

A 12AU7  is  employed  as  a Pierce  crystal 
oscillator,  operating  directly  on  the  chosen 
SSB  frequency  in  the  80  meter  band.  The  sec- 
ond section  of  this  tube  is  used  as  an  isolation 
stage,  with  a tuned  plate  circuit,  Li.  The  out- 
put of  the  oscillator  stage  is  link  coupled  to  a 
90°  r-f  phase-shift  network  wherein  the  audio 
signal  from  the  audio  phasing  network  is  com- 
bined with  the  r-f  signals.  Carrier  balance 
is  accomplished  by  adjustment  of  the  two  1000 
ohm  potentiometers  in  the  r-f  phase  network. 
The  output  of  the  r-f  phasing  network  is  cou- 
pled through  Li  to  a single  6CL6  linear  am- 
plifier which  delivers  a 3 watt  peak  SSB  sig- 
nal on  80  meters. 

A cascade  12AT7  and  a single  6C4  comprise 
the  speech  amplifier  used  to  drive  the  audio 
phase  shift  network.  A small  inter-stage  trans- 
former is  used  ro  provide  the  necessary  180° 
audio  phase  shifr  required  by  the  network.  The 
output  of  the  audio  phasing  network  is  cou- 
pled to  a 12AU7  dual  cathode  follower  which 
provides  the  necessary  low  impedance  circuit 
to  match  the  r-f  phasing  network.  A double- 


AUDIO  PHASE  SHIFT 
NETWORK 
I 


H 


TO 

> 

O 

o 


348  Sideband  Transmission 


HANDBOOK 


S.B.  Exciters  349 


SIMPLE  3-WATT  PHASING  TYPE  SSB  EXCITER 


pole  double-throw  switch  in  the  output  circuit 
of  the  cathode  follower  permits  sideband  se- 
lection. 

A Filter-Type  Exciter  A simple  SSB  filter- 
for  80  and  40  Meters  type  exciter  employ- 
ing the  Collins  me- 
chanical filter  illustrates  many  of  the  basic 
principles  of  sideband  generation.  Such  an  ex- 
citer is  shown  in  figure  39.  The  exciter  is  de- 
signed for  operation  in  the  80  or  40  meter 
phone  bands  and  delivers  sufficient  output  to 
drive  a class  ABi  tetrode  such  as  the  2E26,  807, 
or  6146.  A conversion  crystal  may  be  em- 
ployed, or  a separate  conversion  v-f-o  can  be 
used  as  indicated  on  the  schematic  illustration. 

The  exciter  employs  five  tubes,  exclusive 
of  power  supply.  They  are:  6U8  low  frequency 
oscillator  and  r-f  phase  inverter,  6BA6  i-f  am- 
plifier, 6BA7  high  frequency  mixer,  6AG7 
linear  amplifier,  and  12AU7  speech  amplifier 
and  cathode  follower.  The  heart  of  the  exciter 
is  the  balanced  modulator  employing  two 
1N81  germanium  diodes  and  the  455  kc., 
3500-cycle  bandwidth  mechanical  filter.  The 
input  and  output  circuits  of  the  filter  are  re- 
sonated to  455  kc.  by  means  of  small  padding 

capacitors. 

A series-tuned  Clapp  oscillator  covers  the 
range  of  452  kc.  — 457  kc.  permitting  the 


carrier  frequency  to  be  adjusted  to  the  "20 
decibel”  points  on  the  response  curve  of  the 
filter,  as  shown  in  figure  40.  Proper  r-f  sig- 
nal balance  to  the  diode  modulator  may  be 
obtained  by  adjustment  of  the  padding  capaci- 
tor in  the  cathode  circuit  of  the  triode  section 
of  the  6U8  r-f  tube.  Carrier  balance  is  set  by 
means  of  a 5 OK  potentiometer  placed  across 
the  balanced  modulator. 

One  half  of  a 12AU7  serves  as  a speech  am- 
plifier delivering  sufficient  output  from  a 
high  level  crystal  microphone  to  drive  the  sec- 
ond half  of  the  tube  as  a low  impedance  cath- 
ode follower,  which  is  coupled  to  the  balanced 
modulator.  The  two  1N81  diodes  act  as  an 
electronic  switch,  impressing  a double  side- 
band, suppressed-carrier  signal  upon  the  me- 
chanical filter.  By  the  proper  choice  of  fre- 
quency of  the  beating  oscillator,  the  unwanted 
sideband  may  be  made  to  fall  outside  the  pass- 
band  of  the  mechanical  filter.  Thus  a single 
sideband  suppressed-carrier  signal  appears  at 
the  output  of  the  filter.  The  455  kc.  SSB  sig- 
nal is  amplified  by  a 6BA6  pentode  stage,  and 
is  then  converted  to  a frequency  in  the  80 
meter  or  40  meter  band  by  a 6BA7  mixer 
stage.  Either  a crystal  or  an  external  v-f-o  may 
be  used  for  the  mixing  signal. 

To  reduce  spurious  signals,  a double  tuned 


Figure  40 

THE  ''TWENTY  DB"  CARRIER 
POINTS  ON  THE  FILTER  CURVE 

The  beating  oscillator  should  be  adjusted  so 
that  its  frequency  corresponds  to  the  20  db 
attenuation  points  of  the  mechanical  filter 
passband.  The  corrier  of  the  SSB  signal  is 
thus  attenuated  20  db  in  addition  to  the 
inherent  carrier  attenuation  of  the  balanced 
mixer.  A total  carrier  attenuation  of  50  db 
is  achieved.  Unwanted  sideband  rejection  is 
of  the  same  order. 


Figure  41 

THE  PRODUCT  DETECTOR 

The  above  configuration  resembles  pentagrid 
converter  circuit. 


@ 


LOWER  I 

SIDEBAND  I 

CARRIER 

FREQ. 


FREQUENCY  SPECTRUM  WITH 
COMPLEX  MODULATING  WAVE 


DOUBLE  SIDE-BAND  OUTPUT 
FROM  BALANCED  MODULATOR 
WITH  SINE-WAVE  MODULATION 


Figure  42 

DOUBLE-SIDEBAND 
SUPPRESSED-CARRIER  SIGNAL 

The  envelope  shown  at  B also  is  obtained  on 
the  oscilloscope  when  two  audio  frequencies 
of  the  same  amplitude  are  fed  to  the  input 
of  a single-sideband  transmitter. 


S.B.  Exciters  351 


transformer  is  placed  between  the  mixer  stage 
and  the  6AG7  output  stage.  A maximum  sig- 
nal of  3 watts  may  be  obtained  from  the  6AG7 
linear  amplifier. 

Selection  of  the  upper  or  lower  sideband  is 
accomplished  by  tuning  the  6U8  beating  os- 
cillator across  the  passband  of  the  mechanical 
filter,  as  shown  in  figure  40.  If  the  80  meter 
conversion  oscillator  is  placed  on  the  low  fre- 
quency side  of  the  SSB  signal,  placing  the 
6U8  beating  oscillator  on  the  low  frequency 
side  of  the  passband  of  the  mechanical  filter 
will  produce  the  upper  sideband  on  80  meters. 
When  the  beating  oscillator  is  placed  on  the 
high  frequency  side  of  the  passband  of  the 
mechanical  filter  the  lower  sideband  will  be 
generated  on  80  meters.  If  the  80  meter  con- 
version oscillator  is  placed  on  the  high  fre- 
quency side  of  the  SSB  signal,  the  sidebands 
will  be  reversed  from  the  above.  The  variable 
oscillator  should  be  set  at  approximately  the 
20  db  suppression  point  of  the  passband  of 
the  mechanical  filter  for  best  operation,  as 
shown  in  figure  40.  If  the  oscillator  is  closer 
in  frequency  to  the  filter  passband  than  this, 
carrier  rejection  will  suffer.  If  the  oscillator  is 
moved  farther  away  in  frequency  from  the 
passband,  the  lower  voice  frequencies  will  be 
attenuated,  and  the  SSB  signal  will  sound  high- 
pitched  and  tinny.  A little  practice  in  setting 
the  frequency  of  the  beating  oscillator  while 
monitoring  the  80  meter  SSB  signal  in  the 
station  receiver  will  quickly  acquaint  the  oper- 
ator with  the  proper  frequency  setting  of  the 
beating  oscillator  control  for  transmission  of 
either  sideband. 

If  desired,  an  amplitude  modulated  signal 
with  full  carrier  and  one  sideband  may  be 
transmitted  by  placing  the  6U8  low  frequency 
oscillator  just  inside  either  edge  of  the  pass- 
band  of  the  filter  (designated  "AM  point”, 
figure  40 ) . 

After  the  6U8  oscillator  is  operating  over 
the  proper  frequency  range  it  should  be  possi- 
ble to  tune  the  beating  oscillator  tuning  capac- 
itor across  the  passband  of  the  mechanical  fil- 
ter and  obtain  a reading  on  the  S-meter  of  a 
receiver  tuned  to  the  filter  frequency  and  cou- 
pled to  the  input  grid  of  the  6BA6  i-f  ampli- 
fier tube.  The  two  carrier  balance  controls  of 
the  6U8  phase  inverter  section  should  be  ad- 
justed for  a null  reading  of  the  S-meter  when 
the  oscillator  is  placed  in  the  center  of  the 
filter  passband.  The  6BA6  stage  is  now 
checked  for  operation,  and  transformed  Ti 
aligned  to  the  carrier  frequency.  It  may  be 
necessary  to  unbalance  temporarily  potentio- 


meter #2  of  the  6U8  phase  inverter  in  order 
to  obtain  a sufficiently  strong  signal  for  prop- 
er alignment  of  Ti. 

A conversion  crystal  is  next  plugged  in  the 
6BA7  conversion  oscillator  circuit,  and  the  op- 
eration of  the  oscillator  is  checked  by  moni- 
toring the  crystal  frequency  with  a nearby  re- 
ceiver. The  SSB  "carrier”  produced  by  the  un- 
balance of  potentiometer  #2  should  be  heard 
at  the  proper  sideband  frequency  in  either  the 
80  meter  or  40  meter  band.  The  coupled  cir- 
cuit between  the  6BA7  and  the  6AG7  is  re- 
sonated for  maximum  carrier  voltage  at  the 
grid  of  the  amplifier  stage.  Care  should  be 
taken  that  this  circuit  is  tuned  to  the  sideband 
frequency  and  not  to  the  frequency  of  the  con- 
version oscillator.  Finally,  the  6AG7  stage  is 
tuned  for  maximum  output.  When  these  ad- 
justments have  been  completed,  the  455  kc. 
beating  oscillator  should  be  moved  just  out 
of  the  passband  of  the  mechanical  filter.  The 
80  meter  "carrier”  will  disappear.  If  it  does 
not,  there  is  either  energy  leaking  around  the 
filter,  or  the  amplifier  stages  are  oscillating. 
Careful  attention  to  shielding  (and  neutrali- 
zation) should  cure  this  difficulty. 

Audio  excitation  is  now  applied  to  the  ex- 
citer, and  the  S-meter  of  the  receiver  should 
kick  up  with  speech,  but  the  audio  output  of 
the  receiver  should  be  unintelligible.  As  the 
frequency  of  the  beating  oscillator  is  adjusted 
so  as  to  bring  the  oscillator  frequency  within 
the  passband  of  the  mechanical  filter  the  mo- 
dulation should  become  intelligible.  A single 
sideband  a.m.  signal  is  now  being  generated. 
The  BFO  of  the  receiver  should  now  be  turned 
on,  and  the  beating  oscillator  of  the  exciter 
moved  out  of  the  filter  passband.  When  the 
receiver  is  correctly  tuned,  clean,  crisp  speech 
should  be  heard.  The  oscillator  should  be  set 
at  one  of  the  "20  decibel”  points  of  the  filter 
curve,  as  shown  in  figure  40  and  all  adjust- 
ments trimmed  for  maximum  carrier  suppres- 
sion. 

17-8  Reception  of 

Single  Sideband  Signals 

Single-sideband  signals  may  be  received, 
after  a certain  degree  of  practice  in  the  tech- 
nique, in  a quite  adequate  and  satisfactory 
manner  with  a good  communications  receiver. 
However,  the  receiver  must  have  quite  good 
frequency  stability  both  in  the  high-frequency 

oscillator  and  in  the  beat  oscillator.  For  this 

reason,  receivers  which  use  a crystal-con- 
trolled first  oscillator  are  likely  to  offer  a 


352  Sideband  Transmission 


THE  RADIO 


greater  degree  of  satisfaction  than  the  more 
common  type  which  uses  a self-controlled  os- 
cillator. 

Beat  oscillator  stability  in  most  receivers 
is  usually  quite  adequate,  but  many  receivers 
do  not  have  a sufficient  amplitude  of  beat  os- 
cillator injection  to  allow  reception  of  strong 
SSB  signals  without  distortion.  In  such  re- 
ceivers it  is  necessary  either  to  increase  the 
amount  of  beat-oscillator  injection  into  the 
diode  detector,  or  the  manual  gain  control  of 
the  receiver  must  be  turned  down  quite  low. 

The  tuning  procedure  for  SSB  signals  is  as 
follows:  The  SSB  signals  may  first  be  located 
by  tuning  over  the  band  with  receiver  set 
for  the  reception  of  c-w.;  that  is,  with  the  man- 
ual gain  at  a moderate  level  and  with  the  beat 
oscillator  operating.  By  tuning  in  this  manner 
SSB  signals  may  be  located  when  they  are  far 
below  the  amplitude  of  conventional  AM  sig- 
nals on  the  frequency  band.  Then  after  a sig- 
nal has  been  located,  the  beat  oscillator  should 
be  turned  off  and  the  receiver  put  on  a.v.c. 
Following  this  the  receiver  should  be  tuned 
for  maximum  swing  of  the  S meter  with  modu- 
lation of  the  SSB  signal.  It  will  not  be  possible 
to  understand  the  SSB  signal  at  this  time,  but 
the  receiver  may  be  tuned  for  maximum  deflec- 
tion. Then  the  receiver  is  put  back  on  manual 
gain  control,  the  beat  oscillator  is  turned  on 
again,  the  manual  gain  is  turned  down  until 
the  background  noise  level  is  quite  low,  and 
the  heat  oscillator  control  is  varied  until  the 
signal  sounds  natural. 

The  procedure  in  the  preceding  paragraph 
may  sound  involved,  but  actually  all  the  steps 
except  the  last  one  can  be  done  in  a moment. 
However,  the  last  step  is  the  one  which  will 
requite  some  practice.  In  the  first  place,  it  is 
not  known  in  advance  whether  the  upper  or 
lower  sideband  is  being  transmitted.  So  it  will 
be  best  to  start  tuning  the  beat  oscillator  from 
one  side  of  the  pass  band  of  the  receiver  to 
the  other,  rather  than  starting  with  the  beat 
oscillator  near  the  center  of  the  pass  band  as 
is  normal  for  c-w  reception. 

With  the  beat  oscillator  on  the  wrong  side 
of  the  sideband,  the  speech  will  sound  inverted; 
that  is  to  say  that  low-frequency  mcxlulation 
tones  will  have  a high  pitch  and  high-frequen- 
cy modulation  tones  will  have  a low  pitch — 
and  the  speech  will  be  quite  unintelligible. 
With  the  beat  oscillator  on  the  correct  side  of 
the  sideband  but  too  far  from  the  correct  posi- 
tion, the  speech  will  have  some  intelligibility 
but  the  voice  will  sound  quite  high  pitched. 
Then  as  the  correct  setting  for  the  beat  oscilla- 


tor is  approached  the  voice  will  begin  to  sound 
natural  but  will  have  a background  growl  on 
each  syllable.  At  the  correct  frequency  for  the 
beat  oscillator  the  speech  will  clear  complete- 
ly and  the  voice  will  have  a clean,  crisp  qual- 
ity. It  should  also  be  mentioned  that  there  is 
a narrow  region  of  tuning  of  the  beat  oscillator 
a small  distance  on  the  wrong  side  of  the  side- 
band where  the  voice  will  sound  quite  bassy 
and  difficult  to  understand. 

With  a little  experience  it  will  be  possible 
to  identify  the  sound  associated  with  improper 
settings  of  the  beat-oscillator  control  so  that 
corrections  in  the  setting  of  the  control  can  be 
made.  Note  that  the  main  tuning  control  of  the 
receiver  is  not  changed  after  the  sideband 
once  is  tuned  into  the  pass  band  of  the  re- 
ceiver. All  the  fine  tuning  should  be  done  with 
the  beat  oscillator  control.  Also,  it  is  very  im- 
portant that  the  r-f  gain  control  be  turned  to 
quite  a low  level  during  the  tuning  process. 
Then  after  the  signal  has  been  tuned  properly 
the  r-f  gain  may  be  increased  for  good  signal 
level,  or  until  the  point  is  reached  where  best 
oscillator  injection  becomes  insufficient  and 
the  signal  begins  to  distort. 

Single-Sideband  Greatly  simplified  tuning. 
Receivers  ond  coupled  with  strong  atten- 

Adapters  nation  of  undesired  sig- 

nals, can  be  obtained 
through  the  use  of  a single-sideband  receiver 
or  receiver  adapter.  The  exalted  carrier  prin- 
ciple usually  is  employed  in  such  receivers,  with 
a phase-sensitive  system  sometimes  included 
for  locking  the  local  oscillator  to  the  frequency 
of  the  carrier  of  the  incoming  signal.  In  order 
for  the  locking  system  to  operate,  some  carrier 
must  be  transmitted  along  with  the  SSB  signal. 
Such  receivers  and  adapters  include  a means 
for  selecting  the  upper  or  lower  sideband  by 
the  simple  operation  of  a switch.  For  the  re- 
ception of  a single-sideband  signal  the  switch 
obviously  must  be  placed  in  the  correct  posi- 
tion. But  for  the  reception  of  a conventional 
AM  or  phase-modulated  signal,  either  sideband 
may  be  selected,  allowing  the  sideband  with 
the  least  interference  to  be  used. 

The  Product  Detector  An  unusually  satis- 
factory form  of  de- 
modulator for  SSB  service  is  the  product  de- 
tector, shown  in  one  form  in  figure  41.  This 
circuit  is  preferred  since  it  reduces  intermodu- 
lation products  and  does  not  requite  a large 
local  carrier  voltage,  as  contrasted  to  the  more 
common  diode  envelope  detector.  This  product 
detector  operates  much  in  the  same  manner  as 


HANDBOOK 


S.S.B.  Reception  353 


a multi-grid  mixer  tube.  The  SSB  signal  is 
applied  to  the  control  grid  of  the  tube  and 
the  locally  generated  carrier  is  impressed  upon 
the  other  control  grid.  The  desired  audio  out- 
put signal  is  recovered  across  the  plate  resist- 
ance of  the  demodulator  tube.  Since  the  cath- 
ode current  of  the  tube  is  controlled  by  the 
simultaneous  action  of  the  two  grids,  the  cur- 
rent will  contain  frequencies  equal  to  the  sum 
and  difference  between  the  sideband  signal  and 
the  carrier.  Other  frequencies  are  suppressed  by 
the  low-pass  r-f  filter  in  the  plate  circuit  of  the 
stage,  while  the  audio  frequency  is  recovered 
from  the  i-f  sideband  signal. 

17-9  Double  Sideband 
Transmission 

Many  systems  of  intelligence  transmission 
lie  in  the  region  between  amplitude  modula- 
tion on  the  one  hand  and  single  sideband  sup- 
pressed-carrier  transmission  on  the  other  hand. 
One  system  of  interest  to  the  amateur  is  the 
Synchronous  Communications  System,  popular- 
ly known  as  "double  sideband”  (DSB-)  trans- 
mission, wherein  a suppressed-carrier  double 
sideband  signal  is  transmitted  (figure  42). 
Reception  of  such  a signal  is  possible  by  util- 
izing a local  oscillator  phase-control  system 
which  derives  carrier  phase  information  from 
the  sidebands  alone  and  does  not  require  the 
use  of  any  pilot  carrier. 

The  DSB  T ransmitter  A balanced  modulator 

of  the  type  shown  in 
figure  8 may  be  employed  to  create  a DSB 
signal.  For  higher  operating  levels,  a pair  of 
class-C  type  tetrode  amplifier  tubes  may  be 
screen  modulated  by  a push-pull  audio  system 


and  excited  from  a push-pull  r-f  source.  The 
plates  of  such  a modulator  are  connected  in 
parallel  to  the  tank  circuit,  as  shown  in  figure 
43.  This  DSB  modulator  is  capable  of  1 -kilo- 
watt peak  power  output  at  a plate  potential  of 
4000  volts.  The  circuit  is  self-neutralizing  and 
the  tune-up  process  is  much  the  same  as  with 
any  other  class-C  amplifier  stage.  As  in  the 
case  of  SSB,  the  DSB  signal  may  also  be  gen- 
erated at  a low  level  and  amplified  in  linear 
stages  following  the  modulator. 

Synchronous  A DSB  signal  may  be  re- 
DetecHon  ceived  with  difficulty  on  a 

conventional  receiver,  and 
one  of  the  two  sidebands  may  easily  be  received 
on  a single  sideband  receiver.  For  best  recep- 
tion, however,  a phase-locked  local  oscillator 
and  a synchronous  detector  should  be  em- 
ployed. This  operation  may  be  performed  eith- 
er at  the  frequency  of  reception  or  at  a con- 
venient intermediate  frequency.  A block  dia- 


Figure  44 

BLOCK  DIAGRAM  OF  DSB 
RECEIVING  ADAPTER 


354  Sideband  Transmission 


gram  of  a DSB  synchronous  receiver  is  shown 
in  figure  44.  The  DSB  signal  is  applied  to 
two  detectors  having  their  local  oscillator  con- 
version voltages  in  phase  quadrature  to  each 
other  so  that  the  audio  contributions  of  the 
upper  and  lower  sidebands  reinforce  one  an- 
other. The  in-phase  oscillator  voltage  is  ad- 
justed to  have  the  same  phase  as  the  suppressed 
carrier  of  the  transmitted  signal.  The  I-ampli- 
fier  audio  output,  therefore,  will  contain  the 
demodulated  audio  signal,  while  the  Q-ampli- 
fier  (supplied  with  quadrature  oscillator  volt- 
age) will  produce  no  output  due  to  the  quad- 
rature null.  Any  frequency  change  of  the  local 
oscillator  will  produce  some  audio  output  in 
the  Q-amplifier,  while  the  I-amplifier  is  rela- 
tively unaffected.  The  Q-amplifier  audio  will 
have  the  same  polarity  as  the  I-channel  audio 
for  one  direction  of  oscillator  drift,  and  oppo- 
site polarity  for  oscillator  drift  in  the  opposite 
direction.  The  Q-amplifier  signal  level  is  pro- 
portional to  the  magnitude  of  the  local  oscilla- 
tor phase  angle  error  (the  oscillator  drift)  for 
small  errors.  By  combining  the  I-signal  and. 
the  Q-signal  in  the  audio  phase  discriminator 
a d-c  control  voltage  is  developed  which  auto- 
matically corrects  for  local  oscillator  phase  er- 


rors. The  reactance  tube  therefore  locks  the 
local  oscillator  to  the  correct  phase.  Phase  con- 
trol information  is  derived  entirely  from  the 
sideband  component  of  the  signal  and  the 
carrier  (if  present)  is  not  employed.  Phase 
control  ceases  with  no  modulation  of  the  sig- 
nal and  is  reestablished  with  the  reappearance 
of  modulation. 

Interference  Rejection  Interference  falling 
within  the  passband 
of  the  receiver  can  be  reduced  by  proper  com- 
bination of  the  I-  and  Q-  audio  signals.  Under 
phase  lock  conditions,  the  I-signal  is  com- 
posed of  the  audio  signal  plus  the  undesired 
interference,  whereas  the  Q-signal  contains 
only  the  interference  component.  Phase  can- 
cellation obtained  by  combining  the  two  sig- 
nals will  reduce  the  interference  while  still 
adding  the  desired  information  contained  in 
both  side-bands.  The  degree  of  interference  re- 
jection is  dependent  upon  the  ratio  of  inter- 
ference falling  upon  the  two  sidebands  of  the 
received  signal  and  upon  the  basic  design  of 
the  audio  networks.  A schematic  and  descrip- 
tion of  a complete  DSB  receiving  adapter  is 
shown  in  the  June,  1957  issue  of  CQ  maga- 
zine. 


E A E TECHNI-SHEET 
AIRWOUND  INDUCTORS 


COIL  DIA. 

/NCHES 

TURNS  PER 
INCH 

— 

B & W 

AIR  DUX 

INDUCTANCE 

JJH 

COIL  DIA. 

INCHES 

TURNS  PER 
INCH 

B & W 

AIR  DUX 

INDUCTANCE 

JJH 

1 

2 

4 

3001 

404T 

0. 18 

'i- 

4 

— 

1004 

2.75 

0 

— 

40  6 T 

0.40 

6 

— 

1006 

6.30 

6 

3002 

408T 

0.72 

» 

— 

1008 

11.2 

10 

— 

410T 

1.12 

10 

— 

1010 

17.5 

16 

3003 

416T 

2.90 

16 

1016 

42.5 

32 

3004 

432T 

12.0 

'i- 

4 

— 

1 204 

3.9 

5 

8 

4 

3005 

504r 

0.28 

6 

— 

1 206 

8.8 

6 

— 

506T 

0.62 

8 

_ 

1 208 

15.6 

0 

3006 

508T 

1 . 1 

10 

— 

1210 

24.5 

10 

— 

510T 

1.7 

16 

— 

1216 

63.0 

10 

3007 

510r 

4.4 

4 

— 

1404 

5.2 

32 

3008 

532T 

16.0 

6 

— 

1 406 

11.8 

A 

4 

4 

3009 

604r 

0.39 

8 

— 

1408 

21.0 

6 

— 

6O0r 

0.67 

10 

— 

1410 

33.0 

8 

3010 

606T 

16 

— 

1416 

85.0 

10 

— 

6ior 

2.45 

2 

4 

— 

1604 

8.6 

16 

301  1 

6i6r 

6.40 

6 

— 

1606 

15.0 

32 

301  2 

632T 

26.0 

8 

3900 

1608 

26.5 

4 

3013 

804T 

1.0 

10 

3907-1 

1610 

42.0 

6 

— 

806T 

2.3 

16 

— 

1616 

108.0 

8 

3014 

808T 

4.2 

4 

— 

2004 

10.1 

10 

— 

810T 

6.6 

6 

3905-1 

2006 

23.0 

16 

301  5 

816T 

16.6 

e 

3906-1 

2008 

41.0 

32 

3016 

832T 

68.0 

10 

— 

2010 

106.0 

NOTE: 

3 

4 

— 

2404 

14.0 

CO/L  INDUCTANCE  APPROXIMATELY 
PROPORTIONAL  TO  LENGTH.  I.E.,  POR  IXZ 
INDUCTANCE  VALUE,  TRIM  COIL  TO  1/2  LENGTH. 

6 

— 

2406 

31.5 

6 

— 

2408 

56.0 

10 

— 

2410 

89.0 

CHAPTER  EIGHTEEN 


Transmitter  Design 


The  excellence  of  a transmitter  is  a func- 
tion of  the  design,  and  is  dependent  upon  the 
execution  of  the  design  and  the  proper  choice 
of  components.  This  chapter  deals  with  the 
study  of  transmitter  circuitry  and  of  the  basic 
components  that  go  to  make  up  this  circuitry. 
Modern  components  are  far  from  faultless.  Re- 
sistors have  inductance  and  distributed  ca- 
pacity. Capacitors  have  inductance  and  re- 
sistance, and  inductors  have  resistance  and 
distributed  capacity.  None  of  these  residual 
attributes  show  up  on  circuit  diagrams,  yet 
they  are  as  much  responsible  for  the  success 
or  failure  of  the  transmitter  as  are  the  neces- 
sary and  vital  bits  of  resistance,  capacitance 
and  inductance.  Because  of  these  unwanted 
attributes,  the  job  of  translating  a circuit  on 
paper  into  a working  piece  of  equipment  often 
becomes  an  impossible  task  to  those  individ- 
uals who  disregard  such  important  trivia. 
Rarely  do  circuit  diagrams  show  such  pitfalls 
as  ground  loops  and  residual  inductive  cou- 
pling between  stages.  Parasitic  resonant  cir- 
cuits are  rarely  visible  from  a study  of  the 
schematic.  Too  many  times  radio  equipment  is 
rushed  into  service  before  it  has  been  entirely 
checked.  The  immediate  and  only  too  apparent 
results  of  this  enthusiasm  ate  transmitter  in- 
stability, difficulty  of  neutralization,  t.f.  wan- 
dering all  over  the  equipment,  and  a general 
"touchiness”  of  adjustment.  Hand  in  glove 
with  these  problems  go  the  more  serious  ones 
of  TVI,  key-clicks,  and  parasitics.  By  paying 


attention  to  detail,  with  a good  working  know- 
ledge of  the  limitations  of  the  components, 
and  with  a basic  conception  of  the  actions  of 
ground  currents,  the  average  amateur  will  be 
able  to  build  equipment  that  will  work  "just 
like  the  book  says.” 

The  twin  problems  of  TVI  and  parasitics 
are  an  outgrowth  of  the  major  problem  of  over- 
all circuit  design.  If  close  attention  is  paid 
to  the  cardinal  points  of  circuitry  design,  the 
secondary  problems  of  TVI  and  parasitics 
will  in  themselves  be  solved. 


18-1  Resistors 

The  resistance  of  a conductor  is  a function 
of  the  material,  the  form  the  material  takes, 
the  temperature  of  operation,  and  the  frequen- 
cy of  the  current  passing  through  the  resist- 
ance. In  general,  the  variation  in  resistance 
due  to  temperature  is  directly  proportional  to 
the  temperature  change.  With  most  wire-wound 
resistors,  the  resistance  increases  with  tem- 
perature and  returns  to  its  original  value  when 
the  temperature  drops  to  normal.  So-called 
composition  or  carbon  resistors  have  less 
reliable  temperature/resistance  characteris- 
tics. They  usually  have  a positive  tempera- 
ture coefficient,  but  the  retrace  curve  as  the 
resistor  is  cooled  is  often  erratic,  and  in 


356 


Resistors  357 


DEGREES  CENTIGRADE  DEGREES  CENTIGRADE 


Figure  1 


HEAT  CYCLE  OF  UNCONDITIONED 
COMPOSITION  RESISTORS 


HEAT  CYCLE  OF  CONDITIONED 
COMPOSITION  RESISTORS 


many  cases  the  resistance  does  not  return  to 
its  original  value  after  a heat  cycle.  It  is  for 
this  reason  that  care  must  be  taken  when  sol- 
dering composition  resistors  in  circuits  that 
require  close  control  of  the  resistance  value. 
Matched  resistors  used  in  phase-inverter  serv- 
ice can  be  heated  out  of  tolerance  by  the  act 
of  soldering  them  into  the  circuit.  Long  leads 
should  be  left  on  the  resistors  and  a long- 
nose  pliers  should  grip  the  lead  between  the 
iron  and  the  body  of  the  resistor  to  act  as  a 
heat  block.  General  temperature  character- 
istics of  typical  carbon  resistors  are  shown 
in  figure  1.  The  behavior  of  an  individual  re- 


sistor will  vary  from  these  curves  depending 
upon  the  manufacturer,  the  size  and  wattage 
of  the  resistor,  etc. 

Inductance  of  Every  resistor  because  of  its 
Resistors  physical  size  has  in  addition 

to  its  desired  resistance,  less 
desirable  amounts  of  inductance  and  distrib- 
uted capacitance.  These  quantities  are  illus- 
trated in  figure  2A,  the  general  equivalent  cir- 
cuit of  a resistor.  This  circuit  represents  the 
actual  impedance  network  of  a resistor  at  any 
frequency.  At  a certain  specified  frequency 


Figure  2 


EQUIVALENT  CIRCUIT  OF  A RESISTOR 

I 1 

O— vW 1 Xs  I O 


EQUIVALENT  CIRCUIT  OF  A RESISTOR 
AT  A PARTICULAR  FREQUENCY 


Figure  3 

FREQUENCY  EFFECTS  ON  SAMPLE 
COMPOSITION  RESISTORS 


358  Transmitter  Design 


the  radio 


Figure  4 


CURVES  OF  THE  IMPEDANCE  OF  WIRE- 
WOUND  RESISTORS  AT  RADIO 
FREQUENCIES 


the  impedance  of  the  resistor  may  be  thought 
of  as  a series  reactance  (X,)  as  shown  in  fig- 
ure 2B.  This  reactance  may  be  either  induc- 
tive or  capacitive  depending  upon  whether  the 
residual  inductance  or  the  distributed  capaci- 
tance of  the  resistor  is  the  dominating  factor. 
As  a rule,  skin  effect  tends  to  increase  the 
reactance  with  frequency,  while  the  capacity 
between  turns  of  a wire-wound  resistor,  or  ca- 
pacity between  the  granules  of  a composition 
resistor  tends  to  cause  the  reactance  and  re- 
sistance to  drop  with  frequency.  The  behavior 
of  various  types  of  composition  resistors  over 
a large  frequency  range  is  shown  in  figure  3. 
By  proper  component  design,  non-inductive  re- 
sistors having  a minimum  of  residual  react- 
ance characteristics  may  be  constructed.  Even 
these  have  reactive  effects  that  cannot  be  ig- 
nored at  high  frequencies. 

Wirewound  resistors  act  as  low-Q  inductors 
at  radio  frequencies.  Figure  4 shows  typical 
curves  of  the  high  frequency  characteristics 
of  cylindrical  wirewound  resistors.  In  addition 
to  resistance  variations  wirewound  resistors 
exhibit  both  capacitive  and  inductive  react- 
ance, depending  upon  the  type  of  resistor  and 
the  operating  frequency.  In  fact,  such  resis- 
tors perform  in  a fashion  as  low-Q  r-f  chokes 
below  their  parallel  self-resonant  frequency. 


18-2  Capacitors 

The  inherent  residual  characteristics  of  ca- 
pacitors include  series  resistance,  series  in- 
ductance and  shunt  resistance,  as  shown  in 
figure  5.  The  series  resistance  and  inductance 


R SHUNT 


O 1 I — o 


c L Rseries 


Figure  5 

EQUIVALENT  CIRCUIT  OF  A CAPACITOR 


depend  to  a large  extent  upon  the  physical 
configuration  of  the  capacitor  and  upon  the 
material  of  which  it  is  made.  Of  great  interest 
to  the  amateur  constructor  is  the  series  in- 
ductance of  the  capacitor.  At  a certain  fre- 
quency the  series  inductive  reactance  of  the 
capacitor  and  the  c^acitive  reactance  are 
equal  and  opposite,  and  the  capacitor  is  in 
itself  series  resonant  at  this  frequency.  As 
the  operating  frequency  of  the  circuit  in  which 
the  capacitor  is  used  is  increased  above  the 
series  resonant  frequency,  the  effectiveness 
of  the  capacitor  as  a by-passing  element  de- 
teriorates until  the  unit  is  about  as  effective 
as  a block  of  wood. 

By-Pass  The  usual  forms  of  by-pass  ca- 
Capaeltors  pacitors  have  dielectrics  of  paper, 
mica,  or  ceramic.  For  audio  work, 
and  low  frequency  r-f  work  up  to  perhaps  2 Me. 
or  so,  the  paper  capacitors  are  satisfactory 
as  their  relatively  high  internal  inductance  has 
little  effect  upon  the  proper  operation  of  the 
circuit.  The  actual  amount  of  internal  induct- 
ance will  vary  widely  with  the  manufacturing 
process,  and  some  types  of  paper  capacitors 
have  satisfactory  characteristics  up  to  a fre- 
quency of  5 Me.  or  so. 

When  considering  the  design  of  transmitting 
equipment,  it  must  be  remembered  that  while 
the  transmitter  is  operating  at  some  relatively 
low  frequency  of,  say,  7 Me.,  there  will  be 
harmonic  currents  flowing  through  the  various 
by-pass  capacitors  of  the  order  of  10  to  20 
times  the  operating  frequency.  A capacitor 
that  behaves  properly  at  7 Me.  however,  may 
offer  considerable  impedance  to  the  flow  of 
these  harmonic  currents.  For  minimum  har- 
monic generation  and  radiation,  it  is  obviously 
of  greatest  importance  to  employ  by-pass  ca- 
pacitors having  the  lowest  possible  internal 
inductance. 

Mica  dielectric  capacitors  have  much  less 
internal  inductance  than  do  most  paper  con- 
densers. Figure  6 lists  self-resonant  frequen- 
cies of  various  mica  capacitors  having  various 
lead  lengths.  It  can  be  seen  from  inspection 
of  this  table  that  most  mica  capacitors  be- 
come self-resonant  in  the  12-Mc.  to  50-Mc. 
region.  The  inductive  reactance  they  would 
offer  to  harmonic  currents  of  100  Me.  or  so 


HANDBOOK 


Capacitors  359 


CONDENSER 

LEAD  LENGTHS 

RESONANT  FREQ.  | 

.02  JJfMICA 

NONE 

44.5 

MC. 

.002  JJF  MICA 

NONE 

23.5 

MC. 

.01  UF  MICA 

10 

MC. 

.0009  JUF  MICA 

ir' 

55 

MC. 

.002  JUF  CERAMIC 

r 

24 

MC. 

.001  JJF  CERAMIC 

•ir 

55 

MC. 

500  JJUJF  BUTTON 

NONE 

220 

MC. 

.001  JUF  CERAMIC 

4 

90 

MC. 

.01  JJF  CERAMIC 

i" 

14.5 

MC. 

Figure  6 

SELF-RESONANT  FREQUENCIES  OF 
VARIOUS  CAPACITORS  WITH 
RANDOM  LEAD  LENGTH 


would  be  of  considerable  magnitude.  In  certain 
instances  it  is  possible  to  deliberately  series- 
resonate  a mica  capacitor  to  a certain  frequen- 
cy somewhat  below  its  normal  self-resonant 
frequency  by  trimming  the  leads  to  a critical 
length.  This  is  sometimes  done  for  maximum 
by-passing  effect  in  the  region  of  40  Me.  to 
60  Me. 

The  recently  developed  button-mica  capaci- 
tors shown  in  figure  7 ate  especially  designed 
to  have  extremely  low  internal  inductance. 
Certain  types  of  button-mica  capacitors  of 
small  physical  size  have  a self-resonant  fre- 
quency in  the  region  of  600  Me. 

Ceramic  dielectric  capacitors  in  general 
have  the  lowest  amount  of  series  inductance 
pet  unit  of  capacitance  of  these  three  univer- 
sally used  types  of  by-pass  capacitors.  Typi- 
cal resonant  frequencies  of  various  ceramic 
units  are  listed  in  figure  6.  Ceramic  capaci- 
tors are  available  in  various  voltage  and  capa- 
city ratings  and  different  physical  configura- 
tions. Stand-off  types  such  as  shown  in  figure 
7 are  useful  for  by-passing  socket  and  trans- 
former terminals.  Two  of  these  capacitors  may 
be  mounted  in  close  proximity  on  a chassis 
and  connected  together  by  an  r-f  choke  to  form 
a highly  effective  t-f  filter.  The  inexpensive 
"clamshell”  type  of  ceramic  capacitor  is 
recommended  for  general  by-passing  in  r-f  cir- 
cuitry, as  it  is  effective  as  a by-pass  unit  to 
well  over  100  Me. 

The  large  TV  "doorknob”  capacitors  are 
useful  as  by-pass  units  for  high  voltage  lines. 
These  capacitors  have  a value  of  500  micro- 
microfarads,  and  are  available  in  voltage  rat- 
ings up  to  40,000  volts.  The  dielectric  of 
these  capacitors  is  usually  titanium-dioxide. 
This  material  exhibits  piezo-electric  effects, 

and  capacitors  employing  it  for  a dielectric 
will  tend  to  "talk-back”  when  a-c  voltages 
are  applied  across  them.  When  these  capaci- 


tors are  used  as  plate  bypass  units  in  a modu- 
lated transmitter  they  will  cause  acoustical 
noise.  Otherwise  they  are  excellent  for  gen- 
eral r-f  work. 

A recent  addition  to  the  varied  line  of  ca- 
pacitors is  the  coaxial  or  **Hypass  type  of 
capacitor.  These  capacitors  exhibit  superior 
by-passing  qualities  at  frequencies  up  to  200 
Me.  and  the  bulkhead  type  are  especially  ef- 
fective when  used  to  filter  leads  passing 
through  partition  walls  between  two  stages. 

Variable  Air  Even  though  air  is  the  perfect 
Capacitors  dielectric,  air  capacitors  exhibit 
losses  because  of  the  inherent 
resistance  of  the  metallic  parts  that  make  up 
the  capacitor.  In  addition,  the  leakage  loss 
across  the  insulating  supports  may  become  of 
some  consequence  at  high  frequencies.  Of 
greater  concern  is  the  inductance  of  the  ca- 
pacitor at  high  frequencies.  Since  the  capaci- 
tor must  be  of  finite  size,  it  will  have  tie-rods 
and  metallic  braces  and  end  plates,  all  of  which 
contribute  to  the  inductance  of  the  unit.  The 
actual  amount  of  the  inductance  will  depend 
upon  the  physical  size  of  the  capacitor  and 
the  method  used  to  make  contact  to  the  stator 
and  rotor  plates.  This  inductance  may  be  cut 
to  a minimum  value  by  using  as  small  a ca- 
pacitor as  is  practical,  by  using  insulated  tie- 
rods  to  prevent  the  formation  of  closed  induc- 
tive loops  in  the  frame  of  the  unit,  and  by 
making  connections  to  the  centers  of  the  plate 
assemblies  rather  than  to  the  ends  as  is  com- 


Flgure  7 

TYPES  OF  CERAMIC  AND  MICA  CAPACI- 
TORS SUITABLE  FOR  HIGH-FREQUENCY 
BYPASSING 

The  Centralab  8S8S  (1000  pp/d)  Is  recom- 
mended for  screen  and  plate  circuits  of  tet- 
rode tubes. 


360  Transmitter  Design 


the  radio 


monly  done.  A large  transmitting  capacitor 
may  have  an  inherent  inductance  as  large  as 
0.1  microhenry,  making  the  capacitor  suscep- 
tible to  parasitic  resonances  in  the  50  Me.  to 
150  Me.  range  of  frequencies. 

The  question  of  optimum  C/L  ratio  and  ca- 
pacitor plate  spacing  is  covered  in  Chapter 
Thirteen.  For  all-band  operation  of  a high  power 

stage,  it  is  recommended  that  a capacitor  just 
large  enough  for  40-meter  phone  operation  be 
chosen.  (This  will  have  sufficient  capacitance 
for  phone  operation  on  all  higher  frequency 
bands.)  Then  use  fixed  padding  capacitors  for 
operation  on  80  meters.  Such  padding  capaci- 
tors are  available  in  air,  ceramic,  and  vacuum 
types. 

Specially  designed  variable  capacitors  are 
recommended  for  u-h-f  work;  ordinary  capaci- 
tors often  have  "loops”  in  the  metal  frame 
which  may  resonate  near  the  operating  fre- 
quency. 

Variable  Vacuum  Variable  vacuum  capacitors 
Capacitors  because  of  their  small  phy- 

sical size  have  less  inher- 
ent inductance  pet  unit  of  capacity  than  do 
variable  air  capacitors.  Their  losses  are  ex- 
tremely low,  and  their  dielectric  strength  is 
high.  Because  of  increased  production  the 
cost  of  such  units  is  now  within  the  reach  of 
the  designer  of  amateur  equipment,  and  their 
use  is  highly  recommended  in  high  power  tank 
circuits. 


Wire  and  Inductors 

Any  length  of  wire,  no  matter  how  short, 
has  a certain  value  of  inductance.  This  prop- 
erty is  of  great  help  in  making  coils  and  in- 
ductors, but  may  be  of  great  hindrance  when 
it  is  not  taken  into  account  in  circuit  design 
and  construction.  Connecting  circuit  elements 
(themselves  having  residual  inductance)  to- 
gether with  a conductor  possessing  additional 
inductance  can  often  lead  to  puzzling  difficul- 
ties. A piece  of  no.  10  copper  wire  ten  inches 
long  (a  not  uncommon  length  for  a plate  lead 
in  a transmitter)  can  have  a self-inductance  of 
0.15  microhenries.  This  inductance  and  that 
of  the  plate  tuning  capacitor  together  with  the 
plate-to-ground  capacity  of  the  vacuum  tube 
can  form  a resonant  circuit  which  may  lead  to 
parasitic  oscillations  in  the  v-h-f  regions.  To 
keep  the  self-inductance  at  a minimum,  all  r-f 
carrying  leads  should  be  as  short  as  possible 
and  should  be  made  out  of  as  heavy  material 
as  possible. 

At  the  higher  frequencies,  solid  enamelled 
copper  wire  is  most  efficient  for  t-f  leads. 


Tinned  or  stranded  wire  will  show  greater 
losses  at  these  frequencies.  Tank  coil  and 
tank  capacitor  leads  should  be  of  heavier  wire 
than  other  r-f  leads. 

The  best  type  of  flexible  lead  from  the  en- 
velope of  a tube  to  a terminal  is  thin  copper 
strip,  cut  from  thin  sheet  copper.  Heavy,  rigid 
leads  to  these  terminals  may  crack  the  enve- 
lope glass  when  a tube  heats  or  cools. 

Wires  carrying  only  a.f.  or  d.c.  should  be 
chosen  with  the  voltage  and  current  in  mind. 
Some  of  the  low-filament-voltage  transmitting 
tubes  draw  heavy  current,  and  heavy  wire  must 
be  used  to  avoid  voltage  drop.  The  voltage  is 
low,  and  hence  not  much  insulation  is  required. 
Filament  and  heater  leads  are  usually  twisted 
together.  An  initial  check  should  be  made  on 
the  filament  voltage  of  all  tubes  of  25  watts 
or  more  plate  dissipation  rating.  This  voltage 
should  be  measured  right  at  the  tube  sockets. 
If  it  is  low,  the  filament  transformer  voltage 
should  be  raised.  If  this  is  impossible,  heavier 
or  parallel  wires  should  be  used  for  filament 
leads,  cutting  down  their  length  if  possible. 

Coaxial  cable  may  be  used  for  high  voltage 
leads  when  it  is  desirable  to  shield  them  from 
r-f  fields.  RG-8/U  cable  may  be  used  at  d-c 
potentials  up  to  8000  volts,  and  the  lighter 
RG-17/U  may  be  used  to  potentials  of  3000 
volts.  Spark-plug  type  high-tension  wire  may 
be  used  for  unshielded  leads,  and  will  with- 
stand 10,000  volts. 

If  this  cable  is  used,  the  high-voltage  leads 
may  be  cabled  with  filament  and  other  low- 
voltage  leads.  For  high-voltage  leads  in  low- 
poT/er  exciters,  where  the  plate  voltage  is  not 
over  450  volts,  ordinary  radio  hookup  wire  of 
good  quality  will  serve  the  purpose. 

No  r-f  leads  should  be  cabled;  in  fact  it  is 
better  to  use  enamelled  or  bare  copper  wire 
for  r-f  leads  and  rely  upon  spacing  for  insula- 
tion. All  r-f  joints  should  be  soldered,  and  the 
joint  should  be  a good  mechanical  junction 
before  solder  is  applied. 

The  efficiency  and  Q of  air  coils  commonly 
used  in  amateur  equipment  is  a factor  of  the 
shape  of  the  coil,  the  proximity  of  the  coil  to 
other  objects  (including  the  coil  form)  and  the 
material  of  which  the  coil  is  made.  Dielectric 
losses  in  so-called  "air  wound”  coils  are 
low  and  the  Q of  such  coils  runs  in  the  neigh- 
borhood of  300  to  500  at  medium  frequencies. 
Unfortunately,  most  of  the  transmitting  type 
plug-in  coils  on  the  market  designed  for  link 
coupling  have  far  too  small  a pick  up  link  for 
proper  operation  at  7 Me.  and  3.5  Me.  The  co- 
efficient of  coupling  of  these  coils  is  about 
0.5,  and  additional  means  must  be  employed 
to  provide  satisfactory  coupling  at  these  low 
frequencies.  Additional  inductance  in  series 
with  the  pick  up  link,  the  whole  being  reso- 


HANDBOOK 


Inductors  361 


Rc 
O ' 


L 

-'WiT'— — O 


Rc 

vW — 'TROT'-j— O 

C DISTRIBUTED 

® 


Rc  C 
O— VW 


L 

— o 


© 


Figure  8 

ELECTRICAL  EQUIVALENT  OF  R-F  CHOKE  AT  VARIOUS  FREQUENCIES 


nated  to  the  operating  frequency  will  often 
permit  satisfactory  coupling. 

Coil  Placement  For  best  Q a coil  should  be 
in  the  form  of  a solenoid  with 
length  from  one  to  two  times  the  diameter.  For 
minimum  interstage  coupling,  coils  should  be 
made  as  small  physically  as  is  practicable. 
The  coils  should  then  be  placed  so  that  ad- 
joining coils  are  oriented  for  minimum  mutual 
coupling.  To  determine  if  this  condition  exists, 
apply  the  following  test:  the  axis  of  one  of  the 
two  coils  must  lie  in  the  plane  formed  by  the 
center  turn  of  the  other  coil.  If  this  condition 
is  not  met,  there  will  be  appreciable  coupling 
unless  the  unshielded  coils  are  very  small  in 
diameter  or  are  spaced  a considerable  dis- 
tance from  each  other. 

Insulation  On  frequencies  above  7 Me.,  cera- 
mic, polystyrene,  or  Mycalex  in- 
sulation is  to  be  recommended.  Cold  flow  must 
be  considered  when  using  polystyrene  (Am- 
phenol 912,  etc.).  Bakelite  has  low  losses  on 
the  lower  frequencies  but  should  never  be 
used  in  the  field  of  high-frequency  tank  cir- 
cuits. 

Lucite  (or  Plexiglas),  which  is  available 
in  rods,  sheets,  or  tubing,  is  satisfactory  for 
use  at  all  radio  frequencies  where  the  r-f  volt- 
ages are  not  especially  high.  It  is  very  easy 
to  work  with  ordinary  tools  and  is  not  expen- 
sive. The  loss  factor  depends  to  a consider- 
able extent  upon  the  amount  and  kind  of  plas- 
ticizer used. 

The  most  important  thing  to  keep  in  mind 
regarding  insulation  is  that  the  best  insulation 
is  ait.  If  it  is  necessary  to  reinforce  air-wound 
coils  to  keep  turns  from  vibrating  or  touching, 
use  strips  of  Lucite  or  polystyrene  cemented 
in  place  with  Amphenol  912  coil  dope.  This 
will  result  in  lower  losses  than  the  commonly 
used  celluloid  ribs  and  Duco  cement. 

Radio  Frequency  R-f  chokes  may  be  consid- 
Chokes  ered  to  be  special  induct- 

ances designed  to  have  a 
high  value  of  impedance  over  a large  range  of 
frequencies.  A practical  r-f  choke  has  induct- 
ance, distributed  capacitance,  and  resistance. 


At  low  frequencies,  the  distributed  capacity 
has  little  effect  and  the  electrical  equivalent 
circuit  of  the  r-f  choke  is  as  shown  in  figure 
8A.  As  the  operating  frequency  of  the  choke 
is  raised  the  effect  of  the  distributed  capacity 
becomes  more  evident  until  at  some  particular 
frequency  the  distributed  capacity  resonates 
with  the  inductance  of  the  choke  and  a parallel 
resonant  circuit  is  formed.  This  point  is  shown 
in  figure  8B.  As  the  frequency  of  operation 
is  further  increased  the  overall  reactance  of 
the  choke  becomes  capacitive,  and  finally  a 
point  of  series  resonance  is  reached  (figure 
8C.).  This  cycle  repeats  itself  as  the  operating 
frequency  is  raised  above  the  series  resonant 
point,  the  impedance  of  the  choke  rapidly  be- 
coming lower  on  each  successive  cycle.  A 
chart  of  this  action  is  shown  in  figure  9.  It 
can  be  seen  that  as  the  r-f  choke  approaches 
and  leaves  a condition  of  series  resonance, 
the  performance  of  the  choke  is  seriously  im- 
paired. The  condition  of  series  resonance 
may  easily  be  found  by  shorting  the  terminals 
of  the  t-f  choke  in  question  with  a piece  of 
wire  and  exploring  the  windings  of  the  choke 
with  a grid-dip  oscillator.  Most  commercial 
transmitting  type  chokes  have  series  reso- 
nances in  the  vicinity  of  11  Me.  or  24  Me. 


*2  za. 


CHOKE  IMPEDANCE  (MEGOHM 

— ip  b (K  ; 

On 

20  X-H 

0 

1 

u 

X 

^40 

1 

/i 

A 

t 

k-i5x 

I0X-* 

/ 

^ CHOI 

eB-A 

/ 

A/ 

TO 

/ 

V 

1/ 

7^ 

_ 

I 5 10  15  20  25  30 


FREQUENCY  (MC.) 

Figure  9 

FREQUENCY-IMPEDANCE  CHARACTERIS- 
TICS FOR  TYPICAL  PIE-WOUND 
R-F  CHOKES 


362  Transmitter  Design 


THE  RADIO 


Figure  10 

GROUND  LOOPS  IN  AMPLIFIER  STAGES 

A.  Using  chassis  return 

B.  Common  ground  point 


18-4  Grounds 

At  frequencies  of  30  Me.  and  below,  a chas- 
sis may  be  considered  as  a fixed  ground  refer- 
ence, since  its  dimensions  are  only  a fraction 
of  a wavelength.  As  the  frequency  is  increased 
above  30  Me.,  the  chassis  must  be  considered 
as  a conducting  sheet  on  which  there  are 
points  of  maximum  current  and  potential.  How- 
ever, for  the  lower  amateur  frequencies,  an 
object  may  be  assumed  to  be  at  ground  poten- 
tial when  it  is  affixed  to  the  chassis. 

In  transmitter  stages,  two  important  current 
loops  exist.  One  loop  consists  of  the  grid  cir- 
cuit and  chassis  return,  and  the  other  loop 
consists  of  the  plate  circuit  and  chassis  re- 
turn. These  two  loops  ate  shown  in  figure  10 A. 
It  can  be  seen  that  the  chassis  forms  a return 
for  both  the  grid  and  plate  circuits,  and  that 
ground  currents  flow  in  the  chassis  towards 
the  cathode  circuit  of  the  stage.  For  some 
years  the  theory  has  been  to  separate  these 
ground  currents  from  the  chassis  by  returning 
all  ground  leads  to  one  point,  usually  the 
cathode  of  the  tube  for  the  stage  in  question. 
This  is  well  and  good  if  the  ground  leads  are 
of  minute  length  and  do  not  introduce  cross 
couplings  between  the  leads.  Such  a technique 
is  illustrated  in  figure  lOB.  wherein  all  stage 
components  are  grounded  to  the  cathode  pin 
of  the  stage  socket.  However,  in  transmitter 


construction  the  physical  size  of  the  compon- 
ents prevent  such  close  grouping.  It  is  nec- 
essary to  spread  the  components  of  such  a 
stage  over  a fairly  large  area.  In  this  case  it 
is  best  to  ground  items  directly  to  the  chassis 
at  the  nearest  possible  point,  with  short,  direct 
grounding  leads.  The  ground  currents  will 
flow  from  these  points  through  the  low  induct- 
ance chassis  to  the  cathode  return  of  the 
stage.  Components  grounded  on  the  top  of  the 
chassis  have  their  ground  currents  flow  through 
holes  to  the  cathode  circuit  which  is  usually 
located  on  the  bottom  of  the  chassis,  since 
such  currents  travel  on  the  surface  of  the  chas- 
sis. The  usual  "top  to  bottom”  ground  path 
is  through  the  hole  cut  in  the  chassis  for  the 
tube  socket.  When  the  gain  per  stage  is  rela- 
tively low,  or  there  are  only  a small  number 
of  stages  on  a chassis  this  universal  ground- 
ing system  is  ideal.  It  is  only  in  high  gain 
stages  (i-f  strips)  where  the  "gain  per  inch” 
is  very  high  that  circulating  ground  currents 
will  cause  operational  instability. 

Intercoupling  of  It  is  important  to  prevent  in- 
Ground  Currents  tetcoupling  of  various  differ- 
ent ground  currents  when  the 
chassis  is  used  as  a common  ground  return. 
To  keep  this  intercoupling  at  a minimum,  the 
stage  should  be  completely  shielded.  This 
will  prevent  external  fields  from  generating 
spurious  ground  currents,  and  prevent  the 
ground  currents  of  the  stage  from  upsetting 
the  action  of  nearby  stages.  Since  the  ground 
currents  travel  on  the  surface  of  the  metal, 
the  stage  should  be  enclosed  in  an  electrically 
ti^t  box.  When  this  is  done,  all  ground  cur- 
rents generated  inside  the  box  will  remain  in 
the  box.  The  only  possible  means  of  escape 
for  fundamental  and  harmonic  currents  are  im- 
perfections in  this  electrically  tight  box.  When- 
ever we  bring  a wire  lead  into  the  box,  make 
a ventilation  hole,  or  bring  a control  shaft 
through  the  box  we  create  an  imperfection.  It 
is  important  that  the  effect  of  these  imperfec- 
tions be  reduced  to  a minimum. 


18-5  Holes,  Leads  and  Shafts 

Large  size  holes  for  ventilation  may  be  put 
in  an  electrically  tight  box  provided  they  are 
properly  screened.  Perforated  metal  stock  hav- 
ing many  small,  closely  spaced  holes  is  the 
best  screening  material.  Copper  wire  screen 
may  be  used  provided  the  screen  wires  are 
bonded  together  every  few  inches.  As  the  wire 
corrodes,  an  insulating  film  prevents  contact 
between  the  individual  wires,  and  the  attenua- 
tion of  the  screening  suffers.  The  screening 
material  should  be  carefully  soldered  to  the 


HANDBOOK 


Shielding  363 


box,  or  bolted  with  a spacing  of  not  less  than 
two  inches  between  bolts.  Mating  surfaces  of 
the  box  and  the  screening  should  be  clean. 

A screened  ventilation  opening  should  be 
roughly  three  times  the  size  of  an  equivalent 
unscreened  opening,  since  the  screening  rep- 
resents about  a 70  per  cent  coverage  of  the 
area.  Careful  attention  must  be  paid  to  equip- 
ment heating  when  an  electrically  tight  box 
is  used. 

Commercially  available  panels  having  half- 
inch ventilating  holes  may  be  used  as  part  of 
the  box.  These  holes  have  much  less  attenua- 
tion than  does  screening,  but  will  perform  in  a 
satisfactory  manner  in  all  but  the  areas  of 
weakest  TV  reception.  If  it  is  desired  to  re- 
duce leakage  from  these  panels  to  a minimum, 
the  back  of  the  grille  must  be  coveted  with 
screening  tightly  bonded  to  the  panel. 

Doors  may  be  placed  in  electrically  tight 
boxes  provided  there  is  no  r-f  leakage  around 
the  seams  of  the  door.  Electronic  weather- 
stripping or  metal  "finger  stock”  may  be  used 
to  seal  these  doors.  A long,  narrow  slot  in  a 
closed  box  has  the  tendency  to  act  as  a slot 
antenna  and  harmonic  energy  may  pass  more 
readily  through  such  an  opening  than  it  would 
through  a much  larger  circular  hole. 

Variable  capacitor  shafts  or  switch  shafts 
may  act  as  antennas,  picking  up  currents  in- 
side the  box  and  re-radiating  them  outside  of 
the  box.  It  is  necessary  either  to  ground  the 
shaft  securely  as  it  leaves  the  box,  or  else  to 
make  the  shaft  of  some  insulating  material. 

A two  or  three  inch  panel  meter  requires  a 
large  leakage  hole  if  it  is  mounted  in  the  wall 
of  an  electrically  tight  box.  To  minimize  leak- 
age, the  meter  leads  should  be  by-passed  and 
shielded.  The  meter  should  be  encased  with  a 
metal  shield  that  makes  contact  to  the  box 
entirely  around  the  meter.  The  connecting 
studs  of  the  meter  may  project  through  the 
back  of  the  metal  shield.  Such  a shield  may 
be  made  out  of  the  end  of  a tin  can  of  correct 


RIGHT 


Figure  11B 


Use  of  coaxial  connectors  on  electrically 
tight  box  prevents  escape  of  ground  currents 
from  interior  of  box.  At  the  same  time  exter- 
nal fields  are  not  conducted  into  the  interior 
of  the  box. 


diameter,  cut  to  fit  the  depth  of  the  meter. 
This  complete  shield  assembly  is  shown  in 
figure  11  A. 

Careful  attention  should  be  paid  to  leads 
entering  and  leaving  the  electrically  tight 
box.  Harmonic  currents  generated  within  the 
box  can  easily  flow  out  of  the  box  on  power 
or  control  leads,  or  even  on  the  outer  shields 
of  coaxially  shielded  wires.  Figure  IIB  illus- 
trates the  correct  method  of  bringing  shielded 
cables  into  a box  where  it  is  desired  to  pre- 
serve the  continuity  of  the  shielding. 

Unshielded  leads  entering  the  box  must  be 
carefully  filtered  to  prevent  fundamental  and 
harmonic  energy  from  escaping  down  the  lead. 
Combinations  of  t-f  chokes  and  low  inductance 
by-pass  capacitors  should  be  used  in  power 
leads.  If  the  current  in  the  lead  is  high,  the 
chokes  must  be  wound  of  large  gauge  wire. 
Composition  resistors  may  be  substituted  for 
the  r-f  chokes  in  high  impedance  circuits. 
Bulkhead  or  feed-through  type  capacitors  are 
preferable  when  passing  a lead  through  a 
shield  partition.  A summary  of  lead  le^age 
with  various  filter  arrangements  is  shown  in 
figure  12. 


Internal  Leads  Leads  that  connect  two  points 
within  an  electrically  tight  box 


364  Transmitter  Design 


THE  RADIO 


FIELD  STRENGTH 


TEST  INJUV 

NO. 

1 12000 

2 10000 

Z S30 
4 800 

Z ISO 


^^I^DED  0SCIU-ATOR_ 


B-t-  [ " C 

_LCi 

T 

SHIELDED  HOOK-UP  WIRE 

1 Rp 

L^C2 

^C2 

TCi 

XC3 

TL-Shilld. 

'if' 

Ici  RFC 

_LCi 

^C2 

1 ra  RFC 

-LC4 

Cl  RPC 

T 

-If' 

RFC 

RFC 

-iblyTHRru 

r 

J.C4 

RFC 

^C2  * 



\r-i  RFC  I r-s  SHIELDED'^  WIRE 

L _ _j 


R I - lOOOA  CARBON 
RFC-OHMITE  2-50 


C2  -.005  DISC  CERAMIC 
CS  - .01  SPRAGUE  HI -PASS 
C4  - .005  CERAMIC 
FEED-THROUGH 


Figure  12 

LEAD  LEAKAGE  WITH  VARIOUS 
LEAD  FILTERING  SYSTEMS 
(COURTESY  WIDBM) 


was  operated  near  this  frequency  marked  in- 
stability was  noted,  and  the  filaments  of  the 
810  tubes  increased  in  brilliance  when  plate 
voltage  was  applied  to  the  amplifier,  indicat- 
ing the  presence  of  r.f.  in  the  filament  circuit. 
Changing  the  filament  by-pass  capacitors  to 
■ Ol-pfd.  lowered  the  filament  resonance  fre- 
quency to  2.2  Me.  and  cured  this  effect.  A 
ceramic  capacitor  of  .01-/;tfd.  used  as  a fila- 
ment by-pass  capacitor  on  each  filament  leg 
seems  to  be  satisfactory  from  both  a resonant 
and  a TVl  point  of  view.  Filament  by-pass 
capacitors  smaller  in  value  than  .Ol-pfd. 
should  be  used  with  caution. 

Various  parasitic  resonances  are  also  found 
in  plate  and  grid  tank  circuits.  Push-pull  tank 
circuits  are  prone  to  double  resonances,  as 
shown  in  figure  14.  The  parasitic  resonance 
circuit  is  usually  several  megacycles  higher 
than  the  actual  resonant  frequency  of  the  full 
tank  circuit.  The  cure  for  such  a double  reso- 
nance is  the  inclusion  of  an  t-f  choke  in  the 
center  tap  lead  to  the  split  coil. 

Chassis  Materiol  From  a point  of  view  of  elec- 
trical properties,  aluminum 
is  a poor  chassis  material.  It  is  difficult  to 
make  a soldered  joint  to  it,  and  all  grounds 
must  rely  upon  a pressure  joint.  These  pres- 


may  pick  up  fundamental  and  harmonic  cur- 
rents if  they  ate  located  in  a strong  field  of 
flux.  Any  lead  forming  a closed  loop  with  it- 
self will  pick  up  such  currents,  as  shown  in 
figure  13.  This  effect  is  enhanced  if  the  lead 
happens  to  be  self-resonant  at  the  frequency 
at  which  the  exciting  energy  is  supplied.  The 
solution  for  all  of  this  is  to  by-pass  all  inter- 
nal power  leads  and  control  leads  at  each 
end,  and  to  shield  these  leads  their  entire 
length.  All  filament,  bias,  and  meter  leads 
should  be  so  treated.  This  will  make  the  job 
of  filtering  the  leads  as  they  leave  the  box 
much  easier,  since  normally  "cool”  leads 
within  the  box  will  not  have  picked  up  spur- 
ious currents  from  nearby  "hot”  leads. 


Figure  13 


® 

WRONG 


ILLUSTRATION  OF  HOW  A SUPPOSEDLY 
GROUNDED  POWER  LEAD  CAN  COUPLE 
ENERGY  FROM  ONE  COMPARTMENT 
TO  ANOTHER 


18-6  Parasitic  Resonances 

Filament  leads  within  vacuum  tubes  may 
resonate  with  the  filament  by-pass  capacitors 
at  some  particular  frequency  and  cause  insta- 
bility in  an  amplifier  stage.  Large  tubes  of 
the  810  and  250TH  type  are  prone  to  this  spur- 
ious effect.  In  particular,  a push-pull  810  am- 
plifier using  .001-fifd.  filament  by-pass  capac- 
itors had  a filament  resonant  loop  that  fell  in 
the  7-Mc.  amateur  band.  When  the  amplifier 


ILLUSTRATION  OF  LEAD  ISOLATION  BY 
PROPER  USE  OF  BULKHEAD  BYPASS 
CAPACITOR 


HANDBOOK 


Parasitic  Osciliations  365 


Figure  14 

DOUBLE  RESONANCE  EFFECTS  IN  PUSH- 
PULL  TANK  CIRCUIT  MAY  BE  ELIMI- 
NATED BY  THE  INSERTION  OF  ANY 
R-F  CHOKE  IN  THE  COIL  CENTER 
TAP  LEAD 


sure  joints  are  prone  to  give  trouble  at  a later 
date  because  of  high  resistivity  caused  by 
the  formation  of  oxides  from  eletrolytic  action 
in  the  joint.  However,  the  ease  of  working 
and  forming  the  aluminum  material  far  out- 
weighs the  electrical  shortcomings,  and  alu- 
minum chassis  and  shielding  may  be  used 
with  good  results  provided  care  is  taken  in 
making  all  grounding  connections.  Cadmium 
and  zinc  plated  chassis  ate  preferable  from  a 
corrosion  standpoint,  but  ate  much  more  diffi- 
cult to  handle  in  the  home  workshop. 


18-7  Parasitic  Oscillation 
in  R-F  Amplifiers 

Parasitics  (as  distinguished  from  self-oscil- 
lation on  the  normal  tuned  frequency  of  the 
amplifier)  are  undesirable  oscillations  either 
of  very  high  or  very  low  frequencies  which 
may  occur  in  radio-frequency  amplifiers. 

They  may  cause  spurious  signals  (which 
are  often  rough  in  tone)  other  than  normal  har- 
monics, hash  on  each  side  of  a modulated  car- 
rier, key  clicks,  voltage  breakdown  or  flash- 
over,  instability  or  inefficiency,  and  shortened 
life  or  failure  of  the  tubes.  They  may  be  damped 
and  stop  by  themselves  after  keying  or  modu- 
lation peaks,  or  they  may  be  undamped  and 
build  up  during  ordinary  unmodulated  trans- 
mission, continuing  if  the  excitation  is  re- 
moved. They  may  result  from  series  or  par- 
allel resonant  circuits  of  all  types.  Due  to  neu- 
tralizing lead  length  and  the  nature  of  most 
parasitic  circuits,  the  amplifier  usually  is  not 
neutralized  for  the  parasitic  frequency. 

Sometimes  the  fact  that  the  plate  supply  is 
keyed  will  obscure  parasitic  oscillations  in  a 


final  amplifier  stage  that  might  be  very  severe 
if  the  plate  voltage  were  left  on  and  the  exci- 
tation were  keyed. 

In  some  cases,  an  all-wave  receiver  will 
prove  helpful  in  locating  v-h-f  spurious  oscil- 
lations, but  it  may  be  necessary  to  check  from 
several  hundred  megacycles  downward  in  fre- 
quency to  the  operating  range.  A normal  har- 
monic is  weaker  than  the  fundamental  but  of 
good  tone;  a strong  harmonic  or  a rough  note 
at  any  frequency  generally  indicates  a para- 
sitic. 

In  general,  the  cure  for  parasitic  oscillation 
is  two-fold:  The  oscillatory  circuit  is  damped 
until  sustained  oscillation  is  impossible,  or 
it  is  detuned  until  oscillation  ceases.  An  ex- 
amination of  the  various  types  of  parasitic  os- 
cillations and  of  the  parasitic  oscillatory  cir- 
cuits will  prove  handy  in  applying  the  correct 
cute. 

Low  Frequency  One  type  of  unwanted 

Parasitic  Oscillations  oscillation  often  occurs 
in  shunt-fed  circuits  in 
which  the  grid  and  plate  chokes  resonate,  cou- 
pled through  the  tube’s  interelectrode  capaci- 
tance. This  also  can  happen  with  series  feed. 
This  oscillation  is  generally  at  a much  lower 
frequency  than  the  operating  frequency  and 
will  cause  additional  carriers  to  appear,  spaced 
from  perhaps  twenty  to  a few  hundred  kilo- 
cycles on  either  side  of  the  main  wave.  Such  a 
circuit  is  illustrated  in  figure  15.  In  this  case, 
RFC,  and  RFCj  form  the  grid  and  plate  induct- 
ances of  the  parasitic  oscillator.  The  neutral- 
izing capacitor,  no  longer  providing  out-of- 
phase feedback  to  the  grid  circuit  actually  en- 
hances the  low  frequency  oscillation.  Because 
of  the  low  Q of  the  t-f  chokes,  they  will  usu- 
ally run  warm  when  this  type  of  parasitic  os- 
cillation is  present  and  may  actually  char  and 
burn  up.  A neon  bulb  held  near  the  oscillatory 
circuit  will  glow  a bright  yellow,  the  color 
appearing  near  the  glass  of  the  neon  bulb  and 
not  between  the  electrodes. 

One  cure  for  this  type  of  oscillation  is  to 
change  the  type  of  choke  in  either  the  plate 
or  the  grid  circuit.  This  is  a marginal  cure, 
because  the  amplifier  may  again  break  into  the 
same  type  of  oscillation  when  the  plate  volt- 
age is  raised  slightly.  The  best  cute  is  to  re- 
move the  grid  r-f  choke  entirely  and  replace 
it  with  a wirewound  resistor  of  sufficient  watt- 
age to  catty  the  amplifier  grid  current.  If  the 
inclusion  of  such  a resistor  upsets  the  operat- 
ing bias  of  the  stage,  an  r-f  choke  may  be 
used,  with  a 100-ohm  2-watt  carbon  resistor 
in  series  with  the  choke  to  lower  the  operating 
Q of  the  choke.  If  this  expedient  does  not 
eliminate  the  condition,  and  the  stage  under 

investigation  uses  a beam-tetrode  tube,  nega- 
tive resistance  can  exist  in  the  screen  circuit 


366  Transmitter  Design 


THE  RADIO 


PARASITrc  CIRCUIT  FOR 
LOW  FREQ.  OSCILLATION 


© 


Figure  15 


THE  CAUSE  AND  CURE  OF  LOW  FREQUENCY  PARASITICS 


of  such  tubes.  Try  larger  and  smaller  screen 
by-pass  capacitors  to  determine  whether  or  not 
they  have  any  effect.  If  the  condition  is  com- 
ing from  the  screen  circuit  an  audio  choke  with 
a resistor  across  it  in  series  with  the  screen 
feed  lead  will  often  eliminate  the  trouble. 

Low-frequency  parasitic  oscillations  can 
often  take  place  in  the  audio  system  of  an  AM 
transmitter,  and  their  presence  will  not  be 
known  until  the  transmitter  is  checked  on  a 
receiver.  It  is  easy  to  determine  whether  or 
not  the  oscillations  are  coming  from  the  modu- 
lator simply  by  switching  off  the  modulator 
tubes.  If  the  oscillations  ate  coming  from  the 
modulator,  the  stage  in  which  they  are  being 
generated  can  be  determined  by  removing  tubes 
successively,  starting  with  the  first  speech 
amplification  stage,  until  the  oscillation  stops. 
When  the  stage  has  been  found,  remedial  steps 
can  be  taken  on  that  stage. 

If  the  stage  causing  the  oscillation  is  a low- 
level  speech  stage  it  is  possible  that  the 
trouble  is  coming  from  r-f  or  power-supply 
feedback,  or  it  may  be  coming  about  as  a re- 
sult of  inductive  coupling  between  two  trans- 
formers. If  the  oscillation  is  taking  place  in 
a high-level  audio  stage,  it  is  possible  that 
inductive  or  capacitive  coupling  is  taking 
place  back  to  one  of  the  low-level  speech 
stages.  It  is  also  possible,  in  certain  cases, 
that  parasitic  push-pull  oscillation  can  take 
place  in  a Class  B or  Class  AB  modulator  as 
a result  of  the  grid-to-plate  capacitance  with- 
in the  tubes  and  in  the  stage  wiring.  This  con- 
dition is  mote  likely  to  occur  if  capacitors 
have  been  placed  across  the  secondary  of  the 
driver  transformer  and  across  the  primary  of 
the  modulation  transformer  to  act  in  the  reduc- 
tion of  the  amplitude  of  the  higher  audio  fre- 
quencies. Relocation  of  wiring  or  actual  neu- 
tralization of  the  audio  stage  in  the  manner 
used  for  r-f  stages  may  be  required. 


It  may  be  said  in  general  that  the  presence 
of  low-frequency  parasitics  indicates  that 
somewhere  in  the  oscillating  circuit  there  is 
an  impedance  which  is  high  at  a frequency  in 
the  upper  audio  or  low  r-f  range.  This  imped- 
ance may  include  one  or  more  r-f  chokes  of 
the  conventional  variety,  power  supply  chokes, 
modulation  components,  or  the  high  impedance 
may  be  presented  simply  by  an  RC  circuit 
such  as  might  be  found  in  the  screen-feed  cir- 
cuit of  a beam-tetrode  amplifier  stage. 


18-8  Eliminotion  of  V-H-F 
Parasitic  Oscillations 

V-h-f  parasitic  oscillations  ate  often  diffi- 
cult to  locate  and  difficult  to  eliminate  since 
their  frequency  often  is  only  moderately  above 
the  desired  frequency  of  operation.  But  it  may 
be  said  that  v-h-f  parasitics  always  may  be 
eliminated  if  the  operating  frequency  is  appre- 
ciably below  the  upper  frequency  limit  for  the 
tubes  used  in  the  stage.  However,  the  elimi- 
nation of  a persistent  parasitic  oscillation  on 
a frequency  only  moderately  higher  than  the 
desired  operating  frequency  will  involve  a 
sacrifice  in  either  the  power  output  or  the 
power  sensitivity  of  the  stage,  or  in  both. 

Beam-tetrode  stages,  particularly  those 
using  807  type  tubes,  will  almost  invariably 
have  one  or  more  v-h-f  parasitic  oscillations 
unless  adequate  precautions  have  been  taken 
in  advance.  Many  of  the  units  described  in 
the  constructional  section  of  this  edition  had 
parasitic  oscillations  when  first  constructed. 
But  these  oscillations  were  eliminated  in  each 
case;  hence,  the  e3q>edients  used  in  these 
equipments  should  be  studied.  V-h-f  parasitics 
may  be  readily  identified,  as  they  cause  a 


HANDBOOK 


Parasitic  Oscillations  367 


neon  lamp  to  have  a purple  glow  close  to  the 
electrodes  when  it  is  excited  by  the  parasitic 
energy. 

Parasitic  Oscillations  Triode  stages  are  less 
with  Triodes  subject  to  parasitic  os- 

cillations primarily  be- 
cause of  the  much  lower  power  sensitivity  of 
such  tubes  as  compared  to  beam  tetrodes.  But 
such  oscillations  can  and  do  take  place.  Usual- 
ly, however,  it  is  not  necessary  to  incorporate 
losser  resistors  as  normally  is  the  case  with 
beam  tetrodes,  unless  the  triodes  are  operated 
quite  near  to  their  upper  frequency  limit,  or 
the  tubes  are  characterized  by  a relatively 
high  transconductance.  Triode  v-h-f  parasitic 
oscillations  normally  may  be  eliminated  by  ad- 
justment of  the  lengths  and  effective  induct- 
ance of  the  leads  to  the  elements  of  the  tubes. 

In  the  case  of  triodes,  v-h-f  parasitic  oscil- 
lations often  come  about  as  a result  of  induct- 
ance in  the  neutralizing  leads.  This  is  partic- 
ularly true  in  the  case  of  push-pull  amplifiers. 
The  cure  for  this  effect  will  usually  be  found 
in  reducing  the  length  of  the  neutralizing  leads 
and  increasing  their  diameter.  Both  the  reduc- 
tion in  length  and  increase  in  diameter  will 
reduce  the  inductance  of  the  leads  and  tend 
to  raise  the  parasitic  oscillation  frequency 
until  it  is  out  of  the  range  at  which  the  tubes 
will  oscillate.  The  use  of  straightforward  cir- 
cuit design  with  short  leads  will  assist  in 
forestalling  this  trouble  at  the  outset.  Butter- 
fly-type  tank  capacitors  with  the  neutralizing 
capacitors  built  into  the  unit  (such  as  the 
B&W  type)  ate  effective  in  this  regard. 

V-h-f  parasitic  oscillations  may  take  place 
as  a result  of  inadequate  by-passing  or  long 
by-pass  leads  in  the  filament,  grid-return  and 
plate-return  circuits.  Such  oscillations  also 
can  take  place  when  long  leads  exist  between 
the  grids  and  the  grid  tuning  capacitor  or  be- 
tween the  plates  and  the  plate  tuning  capaci- 
tor. The  grid  and  plate  leads  should  be  kept 
short,  but  the  leads  from  the  tuning  capacitors 
to  the  tank  coils  can  be  of  any  reasonable 
length  insofar  as  parasitic  oscillations  are 
concerned.  In  an  amplifier  where  oscillations 
have  been  traced  to  the  grid  or  plate  leads, 
their  elimination  can  often  be  effected  by  mak- 
ing the  grid  leads  much  longer  than  the  plate 
leads  or  vice  versa.  Sometimes  parasitic  os- 
cillations can  be  eliminated  by  using  iron  or 
nichrome  wire  for  the  grid  or  plate  leads,  or 
for  the  neutralizing  leads.  But  in  any  event 
it  will  always  be  found  best  to  make  the  neu- 
tralizing leads  as  short  and  of  as  heavy  con- 
ductor as  is  practicable. 

In  cases  where  it  has  been  found  that  in- 
creased length  in  the  grid  leads  for  an  ampli- 
fier is  required,  this  increased  length  can  often 
be  wound  into  the  form  of  a small  coil  and  still 


GRID  PARASITIC  SUPPRESSORS  IN  PUSH- 
PULL  TRIODE  STAGE 


obtain  the  desired  effect.  Winding  these  small 
coils  of  iron  or  nichrome  wire  may  sometimes 
be  of  assistance. 

To  increase  losses  at  the  parasitic  frequency, 
the  parasitic  coils  may  be  wound  on  100-ohm 
2-watt  resistots.  These  "lossy”  suppressors 
should  be  placed  in  the  grid  leads  of  the  tubes 
close  to  the  grid  connection,  as  shown  in  fig- 
ure 16. 

Parasitics  with  Where  beam-tetrode  tubes  are 
Beam  Tetrodes  used  in  the  stage  which  has 
been  found  to  be  generating 
the  parasitic  oscillation,  all  the  foregoing 
suggestions  apply  in  general.  However,  there 
are  certain  additional  considerations  involved 
in  elimination  of  parasitics  from  beam-tetrode 
amplifier  stages.  These  considerations  involve 
the  facts  that  a beam-tetrode  amplifier  stage 
has  greater  power  sensitivity  than  an  equiva- 
lent triode  amplifier,  such  a stage  has  a cer- 
tain amount  of  screen-lead  inductance  which 
may  give  rise  to  trouble,  and  such  stages  have 
a small  amount  of  feedback  capacitance. 

Beam-tetrode  stages  often  will  require  the 
inclusion  of  a neutralizing  circuit  to  eliminate 
oscillation  on  the  operating  frequency.  How- 
ever, oscillation  on  the  operating  frequency 
normally  is  not  called  a parasitic  oscillation, 
and  different  measures  are  requited  to  elimi- 
nate the  condition. 

Basically,  parasitic  oscillations  in  beam- 
tetrode  amplifier  stages  fall  into  two  classes: 
cathode-grid-scteen  oscillations,  and  cathode- 
screen-plate  oscillations.  Both  these  types  of 
oscillation  can  be  eliminated  through  the  use 
of  a parasitic  suppressor  in  the  lead  between 
the  screen  terminal  of  the  tube  and  the  screen 
by-pass  suppressor,  as  shown  in  figure  17. 
Such  a suppressor  has  negligible  effect  on  the 
by-passing  effect  of  the  screen  at  the  operat- 
ing frequency.  The  method  of  connecting  this 


THE  RADIO 


368  Transmitter  Design 


Figure  17 

SCREEN  PARASITIC  SUPPRESSION  CIR- 
CUIT FOR  TETRODE  TUBES 


suppressor  to  tubes  having  dual  screen  leads 
is  shown  in  figure  18.  At  the  higher  frequen- 
cies at  which  parasitics  occur,  the  screen  is 
no  longer  at  ground  potential.  It  is  therefore 
necessary  to  include  an  r-f  choke  by-pass  con- 
denser filter  in  the  screen  lead  after  the  para- 
sitic suppressor.  The  screen  lead,  in  addition, 
should  be  shielded  for  best  results. 

During  parasitic  oscillations,  considerable 
r-f  voltage  appears  on  the  screen  of  a tetrode 
tube,  and  the  screen  by-pass  condenser  can 
easily  be  damaged.  It  is  best,  therefore,  to 
employ  screen  by-pass  condensers  whose  d-c 
working  voltage  is  equal  to  twice  the  maximum 
applied  screen  voltage. 

The  grid-screen  oscillations  may  occasion- 
ally be  eliminated  through  the  use  of  a para- 
sitic suppressor  in  series  with  the  grid  lead 
of  the  tube.  The  screen  plate  oscillations  may 
also  be  eliminated  by  inclusion  of  a parasitic 
suppressor  in  series  with  the  plate  lead  of  the 
tube.  A suitable  grid  suppressor  may  be  made 
of  a 22-ohm  2-watr  Ohmite  or  Allen-Bradley 
resistor  wound  with  8 turns  of  no.  18  enameled 
wire.  A plate  circuit  suppressor  is  more  of  a 
problem,  since  it  must  dissipate  a quantity  of 
power  that  is  dependent  upon  just  how  close 
the  parasitic  frequency  is  to  the  operating  fre- 
quency of  the  tube.  If  the  two  frequencies  are 
close,  the  suppressor  will  absorb  some  of  the 
fundamental  plate  circuit  power.  For  kilowatt 
stages  operating  no  higher  than  30  Me.  a satis- 
factory plate  circuit  suppressor  may  be  made 
of  five  570-ohm  2-watt  carbon  resistors  in  par- 
allel, shunted  by  5 turns  of  no.  16  enameled 
wire,  inch  diameter  and  Vi  inch  long  (figure 
19A  and  B). 

The  parasitic  suppressor  for  the  plate  cir- 
cuit of  a small  tube  such  as  the  5763,  2E26, 
807,  6146  or  similar  type  normally  may  con- 
sist of  a 47-ohm  carbon  resistor  of  2-watt  size 
with  6 turns  of  no.  18  enameled  wire  wound 
around  the  resistor.  However,  for  operation 
above  30  Me.,  special  tailoring  of  the  value 


Figure  18 

PHOTO  OF  APPLICATION  OF  SCREEN 
PARASITIC  SUPPRESSION  CIRCUIT 
OF  FIGURE  17 


of  the  resistor  and  the  size  of  the  coil  wound 
around  it  will  be  requited  in  order  to  attain 
satisfactory  parasitic  suppression  without  ex- 
cessive power  loss  in  the  parasitic  suppressor. 

Tetrode  Screening  Isolation  between  the  grid 
and  plate  circuits  of  a tet- 
rode tube  is  not  perfect.  For  maximum  stabili- 
ty, it  is  recommended  that  the  tetrode  stage 
be  neutralized.  Neutralization  is  absolutely 
necessary  unless  the  grid  and  plate  circuits 
of  the  tetrode  stage  are  each  completely  iso- 
lated from  each  other  in  electrically  tight 
boxes.  Even  when  this  is  done,  the  stage  will 
show  signs  of  regeneration  when  the  plate 
and  grid  tank  circuits  are  tuned  to  the  same 
frequency.  Neutralization  will  eliminate  this 
regeneration.  Any  of  the  neutralization  cir- 
cuits described  in  the  chapter  Generation  of 
R-F  Energy  may  be  used. 


18-9  Checking  lor-Parasitic 
Oscillations 

It  is  an  unusual  transmitter  which  harbors 
no  parasitic  oscillations  when  first  constructed 


HANDBOOK 


Parasitic  Oscillations  369 


FOR  807,  ETC. 

PC  = er  #/«£■.  0N47n,zw. 
COMPOSITION  RESISTOR 

FOR  4-250A,  ETC. 
9Z-S'-S70£1.2.W.  COMPOSITION 
RESISTORS  IN  PARALLEL  WITH 
ST.WI6E.  UA"  OIA. 


Figure  19 

PLATE  AND  GRID  PARASITIC  SUPPRESSION  IN  TETRODE  TUBES 


and  tested.  Therefore  it  is  always  wise  to  fol- 
low a definite  procedure  in  checking  a new 
transmitter  for  parasitic  oscillations. 

Parasitic  oscillations  of  all  types  ate  most 
easily  found  when  the  stage  in  question  is 
running  by  itself,  with  full  plate  (and  screen) 
voltage,  sufficient  protective  bias  to  limit  the 
plate  current  to  a safe  value,  and  no  excita- 
tion. One  stage  should  be  tested  at  a time, 
and  the  complete  transmitter  should  never  be 
put  on  the  air  until  all  stages  have  been  thor- 
oughly checked  for  parasitics. 

To  protect  tetrode  tubes  during  tests  for 
parasitics,  the  screen  voltage  should  be  ^- 
plied  through  a series  resistor  which  will  limit 
the  screen  current  to  a safe  value  in  case  the 
plate  voltage  of  the  tetrode  is  suddenly  re- 
moved when  the  screen  supply  is  on.  The  cor- 
rect procedure  for  parasitic  testing  is  as  fol- 
lows (figure  20); 

1.  The  stage  in  question  should  be  coupled 
to  a dummy  load,  and  tuned  up  in  correct  oper- 
ating shape.  Sufficient  protective  bias  should 
be  applied  to  the  tube  at  all  times.  For  pro- 
tection of  the  stage  under  test,  a lamp  bulb 
should  be  added  in  series  with  one  leg  of  the 
primary  circuit  of  the  high  voltage  power  sup- 
ply. As  the  plate  supply  load  increases  during 
a period  of  parasitic  oscillation,  the  voltage 
drop  across  the  lamp  increases,  and  the  effec- 
tive plate  voltage  drops.  Bulbs  of  various 
size  may  be  tried  to  adjust  the  voltage  under 
testing  conditions  to  the  correct  amount.  If  a 
Variac  or  Powerstat  is  at  hand,  it  may  be  used 
in  place  of  the  bulbs  for  smoother  voltage  con- 
trol. Don’t  test  for  parasitics  unless  some  type 
of  voltage  control  is  used  on  the  high  voltage 
supply!  When  a stage  breaks  into  parasitic 
oscillations,  the  plate  current  increases  vio- 
lently, and  some  protection  to  the  tube  under 
test  must  be  used. 

2.  The  t-f  excitation  to  the  tube  should  now 
be  removed.  When  this  is  done,  the  grid,  screen 


and  plate  currents  of  the  tube  should  drop  to 
zero.  Grid  and  plate  tuning  condensers  should 
be  tuned  to  minimum  capacity.  No  change  in 
resting  grid,  screen  or  plate  current  should 
be  observed.  If  a parasitic  is  present,  grid  cur- 
rent will  flow,  and  there  will  be  an  abrupt  in- 
crease in  plate  current.  The  size  of  the  lamp 
bulb  in  series  with  the  high  voltage  supply  may 
be  varied  until  the  stage  can  oscillate  contin- 
uously, without  exceeding  the  rated  plate  dis- 
sipation of  the  tube. 

3.  The  frequency  of  the  parasitic  may  now 
be  determined  by  means  of  an  absorption  wave 
meter,  or  a neon  bulb.  Low  frequency  oscilla- 
tions will  cause  a neon  bulb  to  glow  yellow. 
High  frequency  oscillations  will  cause  the 
bulb  to  have  a soft,  violet  glow.  Once  the  fre- 
quency of  oscillation  is  determined,  the  cures 
suggested  in  this  chapter  may  be  applied  to 
the  stage. 

4.  When  the  stage  can  pass  the  above  test 
with  no  signs  of  parasitics,  the  bias  supply  of 
the  tube  in  question  should  be  decreased  until 
the  tube  is  dissipating  its  full  plate  rating 
when  full  plate  voltage  is  ^plied,  with  no  t-f 


Figure  20 

SUGGESTED  TEST  SETUP  FOR  PARASITIC 
TESTS 


370  Transmitter  Design 


excitation.  Excitation  may  now  be  applied  and 
the  stage  loaded  to  full  input  into  a dummy 
load.  The  signal  should  now  be  monitored  in 
a nearby  receiver  which  has  the  antenna  ter- 
minals grounded  or  otherwise  shorted  out.  A 
seties  of  rapid  dots  should  be  sent,  and  the 
frequency  spectrum  for  several  megacycles 
each  side  of  the  carrier  frequency  carefully 
searched.  If  any  vestige  of  parasitic  is  left, 
it  will  show  up  as  an  occasional  "pop”  on  a 
keyed  dot.  This  "pop”  may  be  enhanced  by  a 
slight  detuning  of  either  the  grid  or  plate  cir- 
cuit. 

5.  If  such  a parasitic  shows  up,  it  means 
that  the  stage  is  still  not  stable,  and  further 
measures  must  be  applied  to  the  circuit.  Para- 
sitic suppressors  maybe  needed  in  both  screen 

and  grid  leads  of  a tetrode,  or  perhaps  in  both 
grid  and  neutralizing  leads  of  a triode  stage. 

As  a last  resort,  a 10,000-ohm  25-watt  wire- 
wound  resistor  may  be  shunted  across  the  grid 
coil,  or  grid  tuning  condenser  of  a high  pow- 
ered stage.  This  strategy  removed  a keying 
pop  that  showed  up  in  a commercial  transmit- 
ter, operating  at  a plate  voltage  of  5000. 


Test  for  Parasitic  It  is  common  experience 

Tendency  in  Tetrode  to  develop  an  engineer- 
Amplifiers  ing  model  of  a new 

equipment  that  is  ap- 
parently free  of  parasitics  and  then  find  trouble- 
some oscillations  showing  up  in  production 
units.  The  reason  for  this  is  that  the  equipment 
has  a parasitic  tendency  that  remains  below 
the  verge  of  oscillation  until  some  change  in 
a component,  tube  gain,  or  operating  condition 
raises  the  gain  of  the  parasitic  circuit  enough 
to  start  oscillation. 

In  most  high  frequency  transmitters  there 
are  a great  many  resonances  in  the  tank  cir- 
cuits at  frequencies  other  than  the  desired 


Figure  21 

PARASITIC  GAIN  MEASUREMENT 
Grid-dip  oscillator  and  vacuum  tube 
voltmeter  may  be  used  to  measure  para- 
sitic stage  gain  over  100kc-20Qmc 
region. 


operating  frequency.  Most  of  these  parasitic 
resonant  circuits  are  not  coupled  to  the  tube 
and  have  no  significant  tendency  to  oscillate. 
A few,  however,  are  coupled  to  the  tube  in 
some  form  of  oscillatory  circuit.  If  the  regener- 
ation is  great  enough,  oscillation  at  the  para- 
sitic frequency  results.  Those  spurious  circuits 
existing  just  below  oscillation  must  be  found 
and  suppressed  to  a safe  level. 

One  test  method  is  to  feed  a signal  from  a 
grid-dip  oscillator  into  the  grid  of  a stage  and 
measure  the  resulting  signal  level  in  the  plate 
circuit  of  the  stage,  as  shown  in  figure  21.  The 
test  is  made  with  all  operating  voltages  applied 
to  the  tubes.  Class  C stages  should  have  bias 
reduced  so  a reasonable  amount  of  static  plate 
current  flows.  The  grid-dip  oscillator  is  tuned 
over  the  range  of  100  kc  to  200  me.  and  the 
relative  level  of  the  r-f  voltmeter  is  watched 
and  the  frequencies  at  which  voltage  peaks 
occur  are  noted.  Each  significant  peak  in  volt- 
age gain  in  the  stage  must  be  investigated.  Cir- 
cuit changes  or  suppression  must  then  be  added 
to  reduce  all  peaks  by  10  db  or  more  in  ampli- 
tude. 


CHAPTER  NINETEEN 


Television  and  Broadcast 
Interference 


The  problem  of  interference  to  television 
reception  is  best  approached  by  the  philoso- 
phy discussed  in  Chapter  Eighteen.  By  correct 
design  procedure,  spurious  harmonic  genera- 
tion in  low  frequency  transmitters  may  be  held 
to  a minimum.  The  remaining  problem  is  two- 
fold: to  make  sure  that  the  residual  harmonics 
generated  by  the  transmitter  are  not  radiated, 
and  to  make  sure  that  the  fundamental  signal 
of  the  transmitter  does  not  overload  the  tele- 
vision receiver  by  reason  of  the  proximity  of 
one  to  the  other. 

In  an  area  of  high  TV-signal  field  intensity 
the  TVI  problem  is  capable  of  complete  solu- 
tion with  routine  measures  both  at  the  amateur 
transmitter  and  at  the  affected  receivers.  But 
in  fringe  areas  of  low  TV-signal  field  srrength 
the  complete  elimination  of  TVI  is  a difficult 
and  challenging  problem.  The  fundamentals 
illustrated  in  Chapter  Fifteen  must  be  closely 
followed,  and  additional  antenna  filtering  of 
rhe  transmitrer  is  required. 

19-1  Types  of  Television 

Interference 

There  are  three  main  types  of  TVI  which 
may  be  caused  singly  or  in  combination  by  the 
emissions  from  an  amateur  transmitter.  These 
types  of  interference  are: 

(1)  Overloading  of  the  TV  set  by  the  trans- 
mitter fundamental 

(2)  Impairment  of  the  picture  by  spurious 

emissions 

(3)  Impairment  of  the  picture  by  the  radia- 
tion of  harmonics 


TV  Set  Even  if  the  amateur  transmitter 

Overloading  were  perfect  and  had  no  har- 
monic radiation  or  spurious 
emissions  whatever,  it  still  would  be  likely  to 
cause  overloading  of  TV  sets  whose  antennas 
were  within  a few  hundred  feet  of  the  trans- 
mitting antenna.  This  type  of  overloading  is 
essentially  the  same  as  the  common  type  of 
BCI  encountered  when  operating  a medium- 
power  or  high-power  amateur  transmitter  with- 
in a few  hundred  feet  of  the  normal  type  of 
BCL  receiver.  The  field  intensity  in  the  im- 
mediate vicinity  of  the  transmitting  antenna 
is  sufficiently  high  that  the  amateur  signal 
will  get  into  the  BC  or  TV  set  either  through 
overloading  of  the  front  end,  or  through  the 
i-f,  video,  or  audio  system.  A characteristic 
of  this  type  of  interference  is  that  it  always 
will  be  eliminated  when  the  transmitter  tem- 
porarily is  operated  into  a dummy  antenna. 
Another  characteristic  of  this  type  of  over- 
loading is  that  its  effects  will  be  substan- 
tially continuous  over  the  entire  frequency 
coverage  of  the  BC  or  TV  receiver.  Channels 
2 through  13  will  be  affected  in  approximately 
the  same  manner. 

With  the  overloading  type  of  interference 
the  problem  is  simply  to  keep  the  fundamental 
of  the  transmitter  out  of  the  affected  receiver. 
Other  types  of  interference  may  or  may  not 
show  up  when  the  fundamental  is  taken  out  of 
the  TV  set  (they  probably  will  appear),  but  at 
least  the  fundamental  must  be  eliminated  first. 

The  elimination  of  the  transmitter  fundamen- 
tal from  the  TV  set  is  normally  the  only  opera- 
tion performed  on  or  in  the  vicinity  of  the  TV 
receiver.  After  the  fundamental  has  been  elimi- 


371 


3 72  TV  and  Broadcast  Interference 


THE  RADIO 


TO  INPUT 
TERMINALS 
OF  TV  SET 


Figure  1 


TUNED  TRAPS  FOR  THE  TRANSMITTER 
FUNDAMENTAL 


The  arrangement  at  (A)  has  proven  to  be  ef- 
fective in  eliminating  the  condition  of  gen- 
eral blocking  as  caused  by  a 28-Mc,  trans- 
mitter in  the  vicinity  of  a TV  receiver.  The 
tuned  circuits  Lj-Ci  are  resonated  separate- 
ly to  the  frequency  of  fransm/ss/on.  The  ad- 
justment may  be  done  at  the  station,  or  it 
may  be  accomplished  at  the  TV  receiver  by 
tuning  for  minimum  interference  on  the  TV 
screen. 


Shown  at  (B)  is  an  alternative  arrangement 
with  a series-tuned  circuit  across  the  anten- 
na terminals  of  the  TV  set.  The  tuned  cir- 
cuit should  be  resonated  to  the  operating 
frequency  of  the  transmitter.  This  arrange- 
ment gives  less  attenuation  of  the  inferfer- 
ing  signal  than  that  at  (A);  the  circuit  has 
proven  effective  with  interference  from  trans- 
mitters on  the  50-Mc.  band,  and  with  low- 
power  28-Mc.  transmitters. 


300  OHM 

LINEFRC^ 

ANTENNA 


SHIELD  BOX 


@ FOR300-OHM  LINE,  SHIELDED  OR  UNSHIELDED 


COAX 

FITTING 


(i)  FOR  50-75  OHM  COAXIAL  LINE 


Figure  2 

HIGH-PASS  TRANSMISSION  LINE  FILTERS 

The  arrangement  at  (A)  will  stop  the  passing 
of  all  signals  below  about  45  Me.  from  the 
antenna  transmission  line  into  the  TV  set. 
Coils  Lj  are  each  1.2  microhenrys  (17  turns 
no.  24  enam.  closewound  on  V4-inch  dia.  poly- 
styrene rod)  with  the  center  tap  grounded. 
It  will  be  found  best  to  scrape,  twist,  and 
solder  the  center  tap  before  winding  the  coil. 
The  number  of  turns  each  side  of  the  tap  may 
then  be  varied  until  the  tap  is  in  the  exact 
center  of  the  winding.  Coil  L2  is  0.6  micro- 
henry (12  turns  no.  24  enam.  closewound  on 
%-inch  dia.  po/ysfyrene  rod).  The  capacitors 
should  be  about  16.5  iJ^fd.,  but  either  15  or 
20  ppfd.  ceramic  copoeftors  will  give  satis- 
factory results.  A similar  filter  for  coaxial 
antenna  fronsm/ss/on  line  is  shown  at  (B). 
Both  coils  should  be  0.12  microhenry  (7  turns 
no.  18  enam.  spaced  to  inch  on  /4-inch  dia, 
polystyrene  rod).  Capacitors  C2  should  be 
75  mifd.  midget  ceramics,  while  C3  should 
be  a 40-mJjfd.  ceramic. 


nated  as  a source  of  interference  to  reception, 
work  may  then  be  begun  on  or  in  the  vicinity 
of  the  transmitter  toward  eliminating  the  other 
two  types  of  interference. 

Taking  Out  More  or  less  standard  BCI- 

the  Fundamental  type  practice  is  most  com- 
monly used  in  taking  out 
fundamental  interference.  Wavetraps  and  fil- 
ters are  installed,  and  the  antenna  system  may 
or  may  not  be  modified  so  as  to  offer  less  re- 
sponse to  the  signal  from  the  amateur  trans- 
mitter. In  regard  to  a comparison  between 
wavetraps  and  filters,  the  same  considerations 
apply  as  have  been  effective  in  regard  to  BCI 
for  many  years;  wavetraps  are  quite  effective 
when  properly  installed  and  adjusted,  but  they 


must  be  readjusted  whenever  the  band  of  oper- 
ation is  changed,  or  even  when  moving  from 
one  extreme  end  of  a band  to  the  other.  Hence, 
wavetraps  are  not  recommended  except  when 
operation  will  be  confined  to  a relatively  nar- 
row portion  of  one  amateur  band.  However,  fig- 
ure 1 shows  two  of  the  most  common  signal 
trapping  arrangements. 

High-Pass  Filters  High-pass  filters  in  the 
antenna  lead  of  the  TV  set 
have  proven  to  be  quite  satisfactory  as  a 
means  of  eliminating  TVI  of  the  overloading 
type.  In  many  cases  when  the  interfering  trans- 
mitter is  operated  only  on  the  bands  below 
7.3  Me.,  the  use  of  a high-pass  filter  in  the 
antenna  lead  has  completely  eliminated  all 


HANDBOOK 


Harmonic  Radiation  373 


TVI.  In  some  cases  the  installation  of  a high- 
pass  filter  in  the  antenna  transmission  line 
and  an  a-c  line  filter  of  a standard  variety  has 
proven  to  be  completely  effective  in  eliminat- 
ing the  interference  from  a transmitter  operat- 
ing in  one  of  the  lower  frequency  amateur 
bands. 

In  general,  it  is  suggested  that  commercial- 
ly manufactured  high-pass  filters  be  purchased. 
Such  units  ate  available  from  a number  of  manu- 
facturers at  a relatively  moderate  cost.  How- 
ever, such  units  may  be  home  constructed; 
suggested  designs  are  given  in  figures  2 and 
3.  Types  for  use  both  with  coaxial  and  with 
balanced  transmission  lines  have  been  shown. 
In  most  cases  the  filters  may  be  constructed 
in  one  of  the  small  shield  boxes  which  are 
now  on  the  market.  Input  and  output  terminals 
may  be  standard  connectors,  or  the  inexpen- 
sive type  of  terminal  strips  usually  used  on 
BC  and  TV  sets  may  be  employed.  Coaxial 
terminals  should  of  course  be  employed  when 
a coaxial  feed  line  is  used  to  the  antenna.  In 
any  event  the  leads  from  the  filter  box  to  the 
TV  set  should  be  very  short,  including  both 
the  antenna  lead  and  the  ground  lead  to  the 
box  itself.  If  the  leads  from  the  box  to  the  set 
have  much  length,  they  may  pick  up  enough 
signal  to  nullify  the  effects  of  the  high-pass 
filter. 

Blocking  from  Operation  on  the  50-Mc.  ama- 
50-Me.  Signals  teur  band  in  an  area  where 
chaimel  2 is  in  use  for  TV 
imposes  a special  problem  in  the  matter  of 
blocking.  The  input  circuits  of  most  TV  sets 
are  sufficiently  broad  so  that  an  amateur  sig- 
nal on  the  50-Mc.  band  will  tide  through  with 
little  attenuation.  Also,  the  normal  TV  antenna 
will  have  a quite  large  response  to  a signal 
in  the  50-Mc.  band  since  the  lower  limit  of 
channel  2 is  54  Me. 

High-pass  filters  of  the  normal  type  simply 
are  not  capable  of  giving  sufficient  attenua- 
tion to  a signal  whose  frequency  is  so  close 
to  the  necessary  pass  bandof  the  filter.  Hence, 
a resonant  circuit  element,  as  illustrated  in 
figure  1,  must  be  used  to  trap  out  the  amateur 
field  at  the  input  of  the  TV  set.  The  trap  must 
be  tuned  or  the  section  of  transmission  line 
cut,  if  a section  of  line  is  to  be  used  for  a 
particular  frequency  in  the  50-Mc.  band. 
This  frequency  will  have  to  be  neat  the  lower 
frequency  limit  of  the  50-Mc.  band  to  obtain 
adequate  rejection  of  the  amateur  signal  while 
still  not  materially  affecting  the  response  of 
the  receiver  to  channel  2. 

Elimination  of  All  spurious  e m i s s i o n s 

Spurious  Emissions  from  amateur  transmitters 
(ignoring  harmonic  signals 
for  the  time  being)  must  be  eliminated  to  com- 


Figure  3 

SERIES-DERIVED  HIGH-PASS  FILTER 

This  filter  is  designed  for  use  in  the 
300-ohm  transmission  line  from  the  TV 
antenna  to  the  TV  receiver,  hlominal  cut- 
off frequency  is  36  Me.  and  maximum  re- 
jection is  at  about  29  Me. 

Cj.Cg — 15-Atrfd.  zero-coefficient  ceramic 

C2,C3,C4,C5 — 20-/i/tfd.  zero-coefficient  cero- 
mic 

L1.L3— 2.0  ^h.  About  24  turns  no.  28  d.c.c. 
wound  to  on  4 " diameter  polystyrene 
rod.  Turns  should  be  adjusted  until  the 
coil  resonates  to  29  Me.  with  the  ossoci- 
oted  iS-mrfi.  copocitor. 

L2 — 0.66  Mb.,  14  turns  no.  28  d.c.c.  wound 
to  on  ^4  dio.  polystyrene  rod.  Adjust 
turns  to  resonate  externally  to  20  Me. 
with  on  auxiliary  100-/i/zfd.  capacitor 
whose  value  is  occurotely  known. 


ply  with  FCC  regulations.  But  in  the  past 
many  amateur  transmitters  have  emitted  spur- 
ious signals  as  a result  of  key  clicks,  para- 
sitics,  and  overmodulation  transients.  In  most 
cases  the  operators  of  the  transmitters  were 
not  aware  of  these  emissions  since  they  were 
radiated  only  for  a short  distance  and  hence 
were  not  brought  to  his  attention.  But  with 
one  or  more  TV  sets  in  the  neighborhood  it 
is  probable  that  such  spurious  signals  will 
be  brought  quickly  to  his  attention. 


^9-2  Harmonic  Radiation 

After  any  condition  of  blocking  at  the  TV 
receiver  has  been  eliminated,  and  when  the 
transmitter  is  completely  free  of  transients 
and  parasitic  oscillations,  it  is  probable  that 
TVI  will  be  eliminated  in  certain  cases.  Cer- 
tainly general  interference  should  be  elimi- 
nated, particularly  if  the  transmitter  is  a well 
designed  affair  operated  on  one  of  the  lower 
frequency  bands,  and  the  station  is  in  a high- 
signal  TV  area.  But  when  the  transmitter  is 
to  be  operated  on  one  of  the  higher  frequency 
bands,  and  particularly  in  a marginal  TV  area, 
the  job  of  TVI-proofing  will  just  have  begun. 
The  elimination  of  harmonic  radiation  from 
the  transmitter  is  a difficult  and  tedious  job 
which  must  be  done  in  an  orderly  manner  if 
completely  satisfactory  results  are  to  be  ob- 
tained. 


374  TV  and  Broadcast  Interference 


THE  RADIO 


TRANSMITTER 

FUNDAMENTAL 

2nd 

3ro 

4 th 

5th 

6th 

7 th 

8th 

9 th 

IOth 

7.0 

21-21.9 

42-44 

56-56.4 

63-65.7 

70-73 

7.3 

TV  I.F. 

NEW 
TV  I F. 

CHANNEL 
© 1 

CHANEL 

CHANEL 

14.0 

42-43 

56-57.6 

70-72 

64-66.4 

98-100.6 

14.4 

NEW 

TV  I.F. 

pH^SEL 

CHANNEL 

@ 

CHANNEL 

© 

FM 

3R0A0- 

CAST 

ro 

o 

63-64.35 

84-65.6 

105-^7.25 

189-193 

210-214.5 

21.  45 

Ctv  i.f.) 

CHi^EL 

CHANNEL 

0 

FM 

BROAD- 

CAST 

CHANN^ 
@ © 

ch^el 

26.96 

27' 23 

53.92- 

54.46 

CHANEL 

AB^  27 
MC  ONLT 

60.66- 

81.69 

CHANEL 

107.64- 

106.92 

FM 

BROAD- 

CAST 

189 

CHANEL 

216 

CH.^NEL 

28.0 

56-59,4 

84-69.1 

166-176.2 

196-207.9 

29.7 

CHANEL 

CHANNEL 

CHANNEL 

@ 

50.0 

100-108 

200-216 

450-486 

500-540 

54.0 

BROAD- 

CAST 

/POSSIBLE  INTERFERENCE 
' TO  u-h-f  channels 

Figure  4 

HARMONICS  OF  THE  AMATEUR  BANDS 


Shown  are  the  harmonic  frequency  ranges  of  the  amateur  bands  between  7 and  S4  Me.,  with  the 
TV  channels  (and  TV  l-f  systems)  which  are  most  likely  to  receive  Interference  from  these  har~ 
monies.  Under  certain  conditions  amateur  signals  In  the  1.8  end  3.S  Me.  bands  can  cause  Inter- 
ference as  a result  of  direct  pickup  In  the  video  systems  of  TV  receivers  which  are  not  ade- 
quately shielded. 


First  it  is  well  to  become  familiar  with  the 
TV  channels  presently  assigned,  with  the  TV 
intermediate  frequencies  commonly  used,  and 
with  the  channels  which  will  receive  inter- 
ference from  harmonics  of  the  various  ama- 
teur bands.  Figures  4 and  5 give  this  infor- 
mation. 

Even  a short  inspection  of  figures  4 and  5 
will  make  obvious  the  seriousness  of  the  in- 
terference which  can  be  caused  by  harmonics 
of  amateur  signals  in  the  higher  frequency 
bands.  With  any  sort  of  reasonable  precautions 
in  the  design  and  shielding  of  the  transmitter 
it  is  not  likely  that  harmonics  higher  than  the 
6th  will  be  encountered.  Hence  the  main  of- 
fenders in  the  way  of  harmonic  interference 
will  be  those  bands  above  14-Mc. 

Nature  of  Investigations  into  the 

Harmonic  Interference  nature  of  the  interfer- 
ence caused  by  ama- 
teur signals  on  the  TV  screen,  assuming  that 
blocking  has  been  eliminated  as  described 
earlier  in  this  chapter,  have  revealed  the  fol- 
lowing facts: 

1.  An  unmodulated  carrier,  such  as  a c-w 
signal  with  the  key  down  or  an  AM  sig- 
nal without  modulation,  will  give  a 


cross-hatch  or  herringbone  pattern  on 
the  TV  screen.  This  same  general  type 
of  picture  also  will  occur  in  the  case  of 
a narrow-band  FM  signal  either  with  or 
without  modulation. 

2.  A relatively  strong  AM  signal  will  give 
in  addition  to  the  herringbone  a very 
serious  succession  of  light  and  dark 
bands  across  the  TV  picture. 

3.  A moderate  strength  c-w  signal  without 
transients,  in  the  absence  of  overload- 
ing of  the  TV  set,  will  result  merely  in 
the  turning  on  and  off  of  the  herringbone 
on  the  picture. 

To  discuss  condition  (1)  above,  the  herring- 
bone is  a result  of  the  beat  note  between  the 
TV  video  carrier  and  the  amateur  harmonic. 
Hence  the  higher  the  beat  note  the  less  ob- 
vious will  be  the  resulting  cross-hatch.  Fur- 
ther, it  has  been  shown  that  a much  stronger 
signal  is  required  to  produce  a discernible 
herringbone  when  the  interfering  harmonic  is 
as  far  away  as  possible  from  the  video  car- 
rier, without  running  into  the  sound  carrier. 
Thus,  as  a last  resort,  or  to  eliminate  the  last 
vestige  of  interference  after  all  corrective 
measures  have  been  taken,  operate  the  trans- 
mitter on  a frequency  such  that  the  intetfet- 


HANDBOOK 


Harmonic  Interference  375 


VIDEO  SOUND 

I I 


f — ' h 

1 h 

— 1 h 

l-H ^ 

— 1 h 

r-t 

r-H h 

1 TV  1 

1 TV  il 

1 TV  1 

1 TV  1 

1 TV  1 

1 TV  1 

1 TV  1 

[CHANNEL  1 

.CHANNEL  1 

.CHANNEL  1 

■ CHANNEL  1 

■ CHANNEL  1 

(CHANNEL  1 

iCHANNEL 1 

1 © 1 
_J L 

1 ® li 

1 ® 1 

1 1 

1 @ 1 

1 II 

1 © 1 
_! !j 

1 © \ 

1 1 

1 © I 

174  180  186  192  198  204  210  216 

HIGH  BAND 


Figure  5 

FREQUENCIES  OF  THE  V-H-F  TV  CHANNELS 


Showing  the  frequency  ranges  of  TV  channels  2 through  13,  with  the  picture  corrier  and  sound 

carrier  frequencies  also  shown* 


ing  harmonic  will  fall  as  far  as  possible  from 
the  picture  carrier.  The  worst  possible  inter- 
ference to  the  picture  from  a continuous  car- 
rier will  be  obtained  when  the  interfering  sig- 
nal is  very  close  in  frequency  to  the  video 
carrier. 

Isolating  Throughout  the  testing  proce- 

the  Source  of  dure  it  will  be  necessary  to 
the  Interference  have  some  sort  of  indicating 
device  as  a means  of  deter- 
mining harmonic  field  intensities.  The  best 
indicator  for  field  intensities  some  distance 
from  the  transmitting  antenna  will  probably  be 
the  TV  receiver  of  some  neighbor  with  whom 
friendly  relations  are  still  maintained.  This 
person  will  then  be  able  to  give  a check,  oc- 
casionally, on  the  relative  nature  of  the  inter- 
ference.-But  it  will  probably  be  necessary  to 
go  and  check  yourself  periodically  the  results 
obtained,  since  the  neighbor  probably  will  not 
be  able  to  give  any  sort  of  a quantitative  anal- 
ysis of  the  progress  which  has  been  made. 

An  additional  device  for  checking  relative- 
ly high  field  intensities  in  the  vicinity  of  the 
transmitter  will  be  almost  a necessity.  A sim- 
ple crystal  diode  wavemetet,  shown  in  figure 
6 will  accomplish  this  function.  Also,  it  will 
be  very  helpful  to  have  a receiver,  with  an  S 
meter,  capable  of  covering  at  least  the  50  to 
100  Me.  range  and  preferably  the  range  to  216 
Me.  This  device  may  consist  merely  of  the 
station  receiver  and  a simple  converter  using 
the  two  halves  of  a 6J6  as  oscillator  and 
mixer. 


The  first  check  can  best  be  made  with  the 
neighbor  who  is  receiving  the  most  serious 
or  the  most  general  interference.  Turn  on  the 
transmitter  and  check  all  channels  to  deter- 
mine the  extent  of  the  interference  and  the 
number  of  channels  affected.  Then  disconnect 
the  antenna  and  substitute  a group  of  100-watt 
lamps  as  a dummy  load  for  the  transmitter.  Ex- 
perience has  shown  that  8 100-watt  lamps  con- 
nected in  two  seriesed  groups  of  four  in  par- 
allel will  take  the  output  of  a kilowatt  trans- 
mitter on  28  Me.  if  connections  are  made  sym- 
metrically to  the  group  of  lamps.  Then  note 
the  interference.  Now  remove  plate  voltage 
from  the  final  amplifier  and  determine  the  ex- 
tent of  interference  caused  by  the  exciter 
stages. 

In  the  average  case,  when  the  final  ampli- 
fier is  a beam  tetrode  stage  and  the  exciter  is 


Crystal-diode  wavemeter  suitable  for  check' 
ing  high-intensity  harmonics  in  TV  region. 


376  TV  and  Broadcast  Interference 


THE  RADIO 


relatively  low  powered  and  adequately  shield- 
ed,  it  will  be  found  that  the  interference  drops 
materially  when  the  antenna  is  removed  and  a 
dummy  load  substituted.  It  will  also  be  found 
in  such  an  average  case  that  the  interference 
will  stop  when  the  exciter  only  is  operating. 

Transmitter  It  should  be  made  clear  at  this 
Power  Level  point  that  the  level  of  power 
used  at  the  transmitter  is  not  of 
great  significance  in  the  basic  harmonic  re- 
duction problem.  The  difference  in  power  level 
between  a 20-watt  transmitter  and  one  rated 
at  a kilowatt  is  only  a matter  of  about  17  db. 
Yet  the  degree  of  harmonic  attenuation  re- 
quired to  eliminate  interference  caused  by 
harmonic  radiation  is  from  80  to  120  db,  de- 
pending upon  the  TV  signal  strength  in  the 
vicinity.  This  is  not  to  say  that  it  is  not  a 
simpler  job  to  eliminate  harmonic  interference 
from  a low-power  transmitter  than  from  a kilo- 
watt equipment.  It  is  simpler  to  suppress  har- 
monic radiation  from  a low-power  transmitter 
simply  because  it  is  a much  easier  problem  to 
shield  a low-power  unit,  and  the  filters  for  the 
leads  which  enter  the  transmitter  enclosure 
may  be  constructed  less  expensively  and 
smaller  for  a low-power  unit. 


19-3  Low-Pass  Filters 


After  the  transmitter  has  been  shielded,  and 
all  power  leads  have  been  filtered  in  such  a 
manner  that  the  transmitter  shielding  has  not 
been  rendered  ineffective,  the  only  remaining 
available  exit  for  harmonic  energy  lies  in  the 
antenna  transmission  line.  Hence  the  main 
burden  of  harmonic  attenuation  will  fall  on  the 
low-pass  filter  installed  between  the  output 
of  the  transmitter  and  the  antenna  system. 

Experience  has  shown  that  the  low-pass 
filter  can  best  be  installed  externally  to  the 
main  transmitter  enclosure,  and  that  the  trans- 
mission line  from  the  transmitter  to  the  low- 
pass  filter  should  be  of  the  coaxial  type. 
Hence  the  majority  of  low-pass  filters  are  de- 
signed for  a characteristic  impedance  of  52 
ohms,  so  that  RG-8/U  cable  (or  RG-58/U  for 
a small  transmitter)  may  be  used  between  the 
output  of  the  transmitter  and  the  antenna  trans- 
mission line  or  the  antenna  tuner. 

Transmitting-type  low-pass  filters  for  ama- 
teur use  usually  are  designed  in  such  a man- 
ner as  to  pass  frequencies  up  to  about  30  Me. 
without  attenuation.  The  nominal  cutoff  fre- 
quency of  the  filters  is  usually  between  38 
and  45  Me.,  and  m-derived  sections  with  maxi- 
mum attenuation  in  chaimel  2 usually  are  in- 
cluded. Well-designed  filters  capable  of  carry- 
ing any  power  level  up  to  one  kilowatt  ate 


r 

L 2^ 

Ls 

L._ 

71 

r i 

CiJL 

i_ 

lC4 

@ 


La 

Ls 

Le 

71 

yUw  V 

TTIRP 

1 

1 

1 

p 

4rC4 

p 

23Af7— ^C3 

T 

TT  T 

- , 

T . 

® 


Figure  7 

LOW-PASS  FILTER  SCHEMATIC  DIAGRAMS 

The  filter  illustrated  at  (A)  uses  m- 
derived  terminating  half  sections  at  each 
end,  with  three  constant^k  mid-sections. 
The  filter  at  (B)  is  essentially  the  same 
except  that  the  center  section  has  been 
changed  to  act  as  an  m-derived  section 
which  can  be  designed  to  offer  maximum 
attenuation  to  channels  2,  4,  5,  or  6 in 
accordance  with  the  constants  given  be- 
low. Cutoff  frequency  is  45  Me.  in  all 
cases.  All  coils,  except  L4  in  (B)  above, 
are  wound  5^  " i,d.  with  8 turns  per  inch. 


The  (A)  Filter 

•— 41.5MMfd.  (40  yU-Mld.  will  be  found  suit- 
able.) 

C2,  C3,C4— 136  MMfd.  (130  to  140  MMfd.  moy  be 
used.) 

Li,Le — 0.2  Mb;  3H  t.  no.  14 
Lj,  L5— 0.3  Mh;  5 t.  no.  12 
L3,  L4,— 0.37  fJ-h;  bVi  t.  no.  12 


The  (B)  Filter  with  Mid-Section  tuned  to  Channel 
2 (58  Me.) 

Cl,  C5— 41.5  /Afd. 

C4~>136  MMfd. 

Cj— 87  MMfd.  (50  /A^fd.  fixed  and  75  /A/ifd.  vari- 
able in  parallel.) 

Li,L7— 0.2Mh;  3ht.  no.  14 
L2,L3,L5,L5— 0.3Mh;  5 t.  no.  12 
L4—O.O9  iiW,  2 t.  no.  14  Vi  " dia.  by  Va  " long 


The  (B)  Filter  with  Mid-Section  tuned  to  Channel 

4 (71  Me.).  All  components  some  except  that; 
C3— .106  MMfd. 

L3,  Lg  —0.33  /iih;  6 t.  no.  12 

L4— 0.05  Mh;  IH  t.  no.  14,  dia.  by  3/8" 

long. 

The  (B)  Filter  with  Mid-Section  tuned  to  Chonnel 

5 (81  Me.).  Change  the  following: 

C3 — 113 

L3,  L5— 0.34  /Ah;  6 t.  no.  12 
L4— 0.033  /Ah;  1 t.  no.  14  3/8  " dia. 


The  (B)  Filter  with  Mid-Section  tuned  to  Channel 
6 (86  Me.).  All  components  are  essentially 
the  same  except  that  the  theoretical  value  of 
L4  is  changed  to  0.03  /Ah.,  and  the  capacitance 
of  C3  is  changed  to  117  /A/Afd. 


HANDBOOK 


Low  Pass  Filters  377 


available  commercially  from  several  manufac- 
turers. Alternatively,  filters  in  kir  form  are 
available  from  several  manufacturers  at  a 
somewhat  lower  price.  Effective  filters  may 
be  home  consrructed,  if  the  test  equipment  is 
available  and  if  sufficient  care  is  taken  in  the 
construction  of  the  assembly. 

Construction  of  Figures  7,  8 and  9 illustrate 
Low-Pass  Filters  high-performance  low-pass 
filters  which  are  suitable 
for  home  construction.  All  are  constructed  in 
slip-cover  aluminum  boxes  (ICA  no.  29110) 
with  dimensions  of  17  by  3 by  2 % inches.  Five 
aluminum  baffle  plates  have  been  installed  in 
the  chassis  to  make  six  shielded  sections 
within  the  enclosure.  Feed-through  bushings 
between  the  shielded  sections  ate  Johnson  no. 
135-55. 

Both  the  (A)  and  (B)  filter  types  ate  de- 
signed for  a nominal  cut-off  frequency  of  45 
Me.,  with  a frequency  of  maximum  rejection 
at  about  57  Me.  as  established  by  the  termi- 
nating half-sections  at  each  end.  Characteris- 
tic impedance  is  52  ohms  in  all  cases.  The 
alternative  filter  designs  diagrammed  in  figure 
7B  have  provision  for  an  additional  rejection 
trap  in  the  center  of  the  filter  unit  which  may 
be  designed  to  offer  maximum  rejection  in 
channel  2,  4,  5,  or  6,  depending  upon  which 
channel  is  likely  to  be  received  in  the  area  in 
question.  The  only  components  which  must  be 
changed  when  changing  the  frequency  of  the 
maximum  rejection  notch  in  the  center  of  the 
filter  unit  are  inductors  L3,  L,,  and  L,,  and 
capacitor  C3.  A trimmer  capacitor  has  been  in- 
cluded as  a portion  of  C3  so  that  the  frequency 
of  maximum  rejection  can  be  tuned  accurately 
to  the  desired  value.  Reference  to  figures  5 
and  6 will  show  the  amateur  bands  which  are 


Figure  8 

PHOTOGRAPH  OF  THE  (B)  FILTER  WITH 
THE  COVER  IN  PLACE 


most  likely  to  cause  interference  to  specific 
TV  channels. 

Either  high-power  or  low-power  components 
maybe  used  in  the  filters  diagrammed  in  figure 
7.  With  the  small  Centralab  TCZ  zero-coeffi- 
cient ceramic  capacirors  used  in  the  filter 
units  of  figure  7A  or  figure  7B,  power  levels 
up  to  200  watts  output  may  be  used  without 
danger  of  damage  to  the  capacitors,  provided 
the  filter  is  feeding  a 52-ohm  resistive  load. 
It  may  be  practicable  to  use  higher  levels  of 
power  with  this  type  of  ceramic  capacitor  in 
the  filter,  but  at  a power  level  of  200  watts  on 
the  28-Mc.  band  the  capacitors  run  just  per- 
ceptibly warm  to  the  touch.  As  a point  of  in- 
terest, it  is  the  current  rating  which  is  of  sig- 
nificance to  the  capacitors  used  in  filters 
such  as  illustrated.  Since  current  ratings  for 
small  capacitors  such  as  these  are  not  readily 
available,  it  is  not  possible  to  establish  an 
accurate  power  rating  for  such  a unit.  The 
high-power  unit  illustrated  in  figure  9,  which 
uses  Centralab  type  850S  and  854S  capacitors. 


Figure  9 

PHOTOGRAPH  OF  THE  (B)  FILTER  WITH  COVER  REMOVED 

The  mid-section  in  this  filter  is  adjusted  for  maximum  rejection  of  channel  4.  Note  that  the  main 
coils  of  the  filter  are  mounted  at  an  angle  of  about  45  degrees  so  that  there  will  be  minimum 
inductive  coupling  from  one  section  to  the  next  through  the  holes  in  the  aluminum  partitions. 
Mounting  the  coils  in  this  manner  was  found  to  give  a measurable  improvement  in  the  attenua- 
tion characteristics  of  the  filter. 


378  TV  and  Broadcast  Interference 


THE  RADIO 


has  proven  quite  suitable  for  power  levels  up 
to  one  kilowatt. 

Capacitors  Ci,  Cj,  C,,  and  C,  can  be  stand- 
ard manufactured  units  with  normal  5 pet  cent 
tolerance.  The  coils  for  the  end  sections  can 
be  wound  to  the  dimensions  given  (Li,  Lj,  and 
L,).  Then  the  resonant  frequency  of  the  series 
resonant  end  sections  should  be  checked  with 
a grid-dip  meter,  after  the  adjacent  input  or 
output  terminal  has  been  shorted  with  a very 
short  lead.  The  coils  should  be  squeezed  or 
spread  until  resonance  occurs  at  57  Me. 

The  intermediate  m-detived  section  in  the 
filter  of  figure  7B  may  also  be  checked  with  a 
grid-dip  meter  for  resonance  at  the  correct  re- 
jection frequency,  after  the  hot  end  of  L,  has 
been  temporarily  grounded  with  a low-induct- 
ance lead.  The  variable  capacitor  portion  of 
C3  can  be  tuned  until  resonance  at  the  correct 
frequency  has  been  obtained.  Note  that  there 
is  so  little  difference  between  the  constants 
of  this  intermediate  section  for  charmels  5 and 
6 that  variation  in  the  setting  of  C3  will  tune 
to  either  chatmel  without  materially  changing 
the  operation  of  the  filter. 

The  coils  in  the  intermediate  sections  of 
the  filter  (Lj,  L3,  L,,  and  L,  in  figure  7A,  and 
Lj,  L3,  Ls , and  Lb  in  figure  7B)  may  be  checked 
most  conveniently  outside  the  filter  unit  with 
the  aid  of  a small  ceramic  capacitor  of  known 
value  and  a grid-dip  meter.  The  ceramic  ca- 
pacitor is  paralleled  across  the  small  coil 
with  the  shortest  possible  leads.  Then  the  as- 
sembly is  placed  atop  a cardboard  box  and  the 
resonant  frequency  checked  with  a grid-dip 
meter.  A Shure  reactance  slide  rule  may  be 
used  to  ascertain  the  correct  resonant  frequen- 
cy for  the  desired  L-C  combination  and  the 
coil  altered  until  the  desired  resonant  frequen- 
cy is  artained.  The  coil  may  then  be  installed 
in  the  filter  unit,  making  sure  that  it  is  not 
squeezed  or  compressed  as  it  is  being  in- 
stalled. However,  if  the  coils  ate  wound  ex- 
actly as  given  under  figure  10,  the  filter  may 
be  assembled  with  reasonable  assurance  that 
it  will  operate  as  designed. 

Using  Low-Pass  The  low-pass  filter  con- 
Filters  nected  in  the  output  trans- 

mission line  of  the  trans- 
mitter is  capable  of  affording  an  enormous  de- 
gree of  harmonic  attenuation.  However,  the 
filter  must  be  operated  in  the  correct  manner 
or  the  results  obtained  will  not  be  up  to  ex- 
pectations. 

In  the  first  place,  all  direct  radiation  from 
the  transmitter  and  its  control  and  power  leads 
must  be  suppressed.  This  subject  has  been 
discussed  in  the  previous  section.  Secondly, 
the  filter  must  be  operated  into  a load  imped- 
ance approximately  equal  to  its  design  char- 
acteristic impedance.  The  filter  itself  will 


SCHEMATIC  OF  THE  SINGLE-SECTION 
HALF-WAVE  FILTER 

The  constants  given  below  are  for  a char- 
acteristic impedance  of  52  ohms,  for  use  with 
RG-8/U  and  RG-58/U  cable.  Coil  L/  should 
be  checked  for  resonance  at  the  operating 
frequency  with  C/,  and  the  same  with  Lj  and 
C^.  This  check  can  be  made  by  sol  dering  a 
low-inductance  grounding  strap  to  the  lead 
between  L/  and  Lj  where  it  passes  through 
the  shield.  When  the  coils  have  been  trimmed 
to  resonance  with  a grid-dip  meter,  the 
grounding  strap  should  of  course  be  removed. 
This  filter  type  will  give  an  attenuation  of 
about  30  db  to  the  second  harmonic,  about 
48  db  to  the  third,  about  60  db  to  the  fourth, 
67  to  the  fifth,  and  so  on  increasing  at  a 
rate  of  about  30  db  per  octave. 

Ci,C2,C3,C^-.—Siiver  mico  or  smoll  ceramic  for  low 
power,  transmitting  type  ceramic  lor  high  power. 
Copocitonce  for  different  bonds  is  given  below: 

160  meters— 1700  fi;Ufd. 

80  meters— 850 

40  meters — 440  p-p-ld. 

20  meters— 220  {xpfi. 

10  meters— 110  p-ilH. 

6 meters— 60  ilflfi. 

Li.Lj— May  be  mode  up  of  sections  of  B8<W  Mini- 
ductor  for  power  levels  below  250  wotts,  or  of 
no.  12  enom.  for  power  up  to  one  kilowatt.  Ap- 
proximate dimensions  for  the  coils  ore  given 
below,  but  the  coils  should  be  trimmed  to  reso- 
nate at  the  proper  frequency  with  a grid-dip  me- 
ter os  discussed  above.  All  coils  except  the 
ones  for  160  meters  ore  wound  8 turns  per  inch. 
160  meters— 4.2  p-h;  22  turns  no.  16  enom..  1 '' 
die.  2"  long 

80  meters— 2.1  phi  13  t.  1 “ dio.  (No.  3014  Mini- 
ductor  or  no.  12) 

40  meters — 1.1  pW,  8 t.  1"  dlo.  (No.  3014  or  no. 
12  at  8 t.p.i.) 

20  motors— 0.55  phi  7 t.*«"  dio.  (No.  3010  or 
no.  12  at  8 t.p.i.) 

10  meters^— 0.3  f^h;  6 t.  dlo.  (No.  3002  or 
no.  12  at  8 t.p.i.) 

6 meters— 0.17  Mh;  4 t.  Vt"  dio.  (No.  3002  or 
no.  12  at  8 t.p.i.) 


have  very  low  losses  (usually  less  than  0.5 
db)  when  operated  into  its  nominal  value  of  re- 
sistive load.  But  if  the  filter  is  mis-terminated 
its  losses  will  become  excessive,  and  it  will 
not  present  the  correct  value  of  load  imped- 
ance to  the  transmitter. 

If  a filter,  being  fed  from  a high-power 
transmitter,  is  operated  into  an  incorrect  ter- 
mination it  may  be  damaged;  the  coils  may  be 
overheated  and  the  capacitors  destroyed  as  a 
result  of  excessive  r-f  currents.  Hence  it  is 
wise,  when  first  installing  a low-pass  filter. 


HANDBOOK 


Broadcast  Interference  379 


Figure  11 

HALF-WAVE  FILTER 
FOR  THE  28.MC.  BAND 

Showing  one  possible  type 
of  construction  of  a 52-ohm 
half-wave  filter  for  relative- 
ly low  power  operation  on 
the  28-Mc.  band. 


to  check  the  standing-wave  ratio  of  the  load 
being  presented  to  the  output  of  the  filter  with 
a standing-wave  meter  of  any  of  the  conven- 
tional types.  Then  the  antenna  termination  or 
the  antenna  coupled  should  be  adjusted,  with 
low  power  on  the  transmitter,  until  the  s.w.r. 
of  the  load  being  presented  to  the  filter  is 
less  than  2.0,  and  preferably  below  1.5. 

Half-Wave  Filters  Half-wave  filters  {”Har- 
monikers”)  have  been  dis- 
cussed in  various  publications  including  the 
Nov. -Dec.  1949  GE  Ham  News.  Such  fUters 
are  relatively  simple  and  offer  the  advantage 
that  they  present  the  same  value  of  impedance 
at  their  input  terminals  as  appears  as  load 
across  their  output  terminals.  Such  filters  nor- 
mally are  used  as  one-band  affairs,  and  they 
offer  high  attenuation  only  to  the  third  and 
higher  harmonics.  Design  data  on  the  half- 
wave filter  is  given  in  figure  10.  Construction 
of  half-wave  filters  is  illustrated  in  figure  11. 


19-4  Broadcast 

Interference 

Interference  to  the  reception  of  signals  in 
the  broadcast  band  (540  to  1600  kc.)  or  in  the 
FM  broadcast  band  (88  to  108  Me.)  by  amateur 
transmissions  is  a serious  matter  to  those 
amateurs  living  in  densely  populated  areas. 
Although  broadcast  interference  has  recently 
been  overshadowed  by  the  seriousness  of  tele- 
vision interference,  the  condition  of  BCI  is 
still  present. 

In  general,  signals  from  a transmitter  oper- 
ating properly  are  not  picked  up  by  receivers 
tuned  to  other  frequencies  unless  the  receiver 
is  of  inferior  design,  or  is  in  poor  condition. 
Therefore,  if  the  receiver  is  of  good  design 


and  is  in  good  repair,  the  burden  of  rectifying 
the  trouble  rests  with  the  owner  of  the  inter- 
fering station.  Phone  and  c-w  stations  both 
are  capable  of  causing  broadcast  interference, 
key-click  annoyance  from  the  code  transmitters 
being  particularly  objectionable. 

A knowledge  of  each  of  the  several  types 
of  broadcast  interference,  their  cause,  and 
methods  of  eliminating  them  is  necessary  for 
the  successful  disposition  of  this  trouble.  An 
effective  method  of  combating  one  variety  of 
interference  is  often  of  no  value  whatever  in 
the  correction  of  another  type.  Broadcast  in- 
terference seldom  can  be  cured  by  "rule  of 
thumb”  procedure. 

Broadcast  interference,  as  covered  in  this 
section  refers  primarily  to  standard  (amplitude 
modulated,  550-1600  kc.)  broadcast.  Interfer- 
ence with  FM  broadcast  reception  is  much 
less  common,  due  to  the  wide  separation  in 
frequency  between  the  FM  broadcast  band  and 
the  mote  popular  amateur  bands,  and  due  also 
to  the  limiting  action  which  exists  in  all  types 
of  FM  receivers.  Occasional  interference  with 
FM  broadcast  by  a harmonic  of  an  amateur 
transmitter  has  been  reported;  if  this  condi- 
tion is  encountered,  it  may  be  eliminated  by 
the  procedures  discussed  in  the  first  portion 
of  this  chapter  under  Television  Interference. 

The  use  of  frequency-modulation  transmis- 
sion by  an  amateur  station  is  likely  to  result 
in  much  less  interference  to  broadcast  recep- 
tion than  either  amplitude-modulated  telephony 
or  straight  keyed  c.w.  This  is  true  because, 
insofar  as  the  broadcast  receiver  is  concerned, 
the  amateur  FM  transmission  will  consist  of  a 
plain  unmodulated  carrier.  There  will  be  no 
key  clicks  or  voice  reception  picked  up  by 
the  b-c-1  set  (unless  it  happens  to  be  an  FM 
receiver  which  might  pick  up  a harmonic  of 
the  signal),  although  there  might  be  a slight 
click  when  the  transmitter  is  put  on  or  taken 


38  0 TV  and  Broadcast  Interference 


THE  RADIO 


WAVE-TRAP  CIRCUITS 


The  circuif  of  (A)  is  the  most  common  ar~ 
rangement,  but  the  circuit  at  (B)  may  give 
improved  results  under  certain  conditians. 
Manufactured  wave  traps  for  the  desired  band 
of  operation  may  be  purchased  or  the  traps 
may  be  assembled  from  the  data  given  In 
figure  14. 


off  the  air.  This  is  one  reason  why  narrow- 
band  FM  has  become  so  popular  with  phone 
enthusiasts  who  reside  in  densely  populated 
areas. 

Interference  Depending  upon  whether  it  is 

Classifications  traceable  directly  to  causes 

within  the  station  or  within 
the  receiver,  broadcast  interference  may  be 
divided  into  two  main  classes.  For  example, 
that  type  of  interference  due  to  transmitter 
over- modulation  is  at  once  listed  as  being 
caused  by  improper  operation,  while  an  inter- 
fering signal  that  tunes  in  and  out  with  a 
broadcast  station  is  probably  an  indication  of 
cross  modulation  or  image  response  in  the  re- 
ceiver, and  the  poorly-designed  input  stage 
of  the  receiver  is  held  liable.  The  various 
types  of  interference  and  recommended  cures 
will  be  discussed  in  the  following  paragraphs. 

Blanketing  This  is  not  a tunable  effect,  but 
a total  blocking  of  the  receiver. 
A more  or  less  complete  "washout”  covets 
the  entire  receiver  range  when  the  carrier  is 
switched  on.  This  produces  either  a complete 
blotting  out  of  all  broadcast  stations,  or  else 
knocks  down  their  volume  several  decibels— 
depending  upon  the  severity  of  the  interfer- 
ence. Voice  modulation  of  the  carrier  causing 
the  blanketing  will  be  highly  distorted  or  even 


Figure  13 

HIGH-ATTENUATION  WAVE-TRAP 
CIRCUIT 

The  two  circuits  may  be  tuned  to  the  same 
frequency  for  highest  attenuation  of  a strong 
signal,  or  the  two  traps  may  be  tuned  sep- 
arately for  different  bands  of  operation. 


unintelligible.  Keying  of  the  carrier  which 
produces  the  blanketing  will  cause  an  annoy- 
ing fluctuation  in  the  volume  of  the  broadcast 
signals. 

Blanketing  generally  occurs  in  the  imme- 
diate neighborhood  (inductive  field)  of  a pow- 
erful transmitter,  the  affected  area  being  di- 
rectly proportional  to  the  power  of  the  trans- 
mitter. Also  it  is  more  prevalent  with  trans- 
mitters which  operate  in  the  160-meter  and 
80-meter  bands,  as  compared  to  those  on  the 
higher  frequencies. 

The  remedies  are  to  (1)  shorten  the  receiv- 
ing antenna  and  thereby  shift  its  resonant  fre- 
quency, or  (2)  remove  it  to  the  interior  of  the 
building,  (3)  change  the  direction  of  either 
the  receiving  or  transmitting  antenna  to  mini- 
mize their  mutual  coupling,  or  (4)  keep  the 
interfering  signal  from  entering  the  receiver 
input  circuit  by  installing  a wavetrap  tuned 
to  the  signal  frequency  (see  figure  12)  or  a 
low-pass  filter  as  shown  in  figure  21. 

A suitable  wave-trap  is  quite  simple  in  con- 
struction, consisting  only  of  a coil  and  midget 
variable  capacitor.  When  the  trap  circuit  is 
tuned  to  the  frequency  of  the  interfering  sig- 
nal, little  of  the  interfering  voltage  reaches 
the  grid  of  the  first  tube.  Commercially  manu- 
factured wave-traps  are  available  from  several 
concerns,  including  the  J.  W.  Miller  Co.  in 
Los  Angeles.  However,  the  majority  of  ama- 
teurs prefer  to  construct  the  traps  from  spate 
components  selected  from  the  "junk  box.” 

The  circuit  shown  in  figure  13  is  particu- 
larly effective  because  it  consists  of  two 
traps.  The  shunt  trap  blocks  or  rejects  the 
frequency  to  which  it  is  tuned,  while  the 
series  trap  across  the  antenna  and  ground  ter- 
minals of  the  receiver  provides  a very  low  im- 
pedance path  to  ground  at  the  frequency  to 


HANDBOOK 


Wavetraps  381 


BAND 

COIL  L 

CAPACITOR,  C 

1.8 

Me. 

1 inch  no.  30  enam. 
closewound  on  1"  form 

75-/t#(fd. 

var. 

3.5 

Me. 

42  turns  no.  30 
closewound  on  1 

enom. 

' form 

50-p/:fd. 

vor. 

7.0 

Me. 

23  turns  no.  24 
closewound  on  1 

enam. 

' form 

50-;.(/ffd. 

Yor. 

14 

Me. 

10  turns  no.  24 
closewound  on  1 

enam. 

' form 

SO-M/ifd. 

var. 

21 

Me. 

7 turns  no.  24 
closewound  on  1 

enam. 

' form 

50-/.f/ifd. 

var. 

28 

Me. 

4 turns  no.  24 
closewound  on  1 

enam. 

' form 

25-A/xfd. 

var. 

50 

Me. 

3 turns  no.  24 
spaced  V2"  on  1 

enam. 

' form 

is-tLfiid. 

var. 

Figure  14 

COIL  AND  CAPACITOR  TABLE  FOR 
AMATEUR-BAND  WAVETRAPS 


Figure  15 

MODIFICATION  OF  THE  FIGURE  13 
CIRCUIT 


In  this  circuit  arrangement  the  parallel-tuned 
tank  is  inductively  coupled  to  the  antenna 
lead  with  a 3 to  6 turn  link  instead  of  being 
placed  directly  in  series  with  the  antenna 
lead^ 


which  it  is  tuned  and  by-passes  the  signal  to 
ground.  In  moderate  interference  cases,  either 
the  shunt  or  series  trap  may  be  used  alone, 
while  similarly,  one  trap  may  be  tuned  to  one 
of  the  frequencies  of  the  interfering  trans- 
mitter and  the  other  trap  to  a different  inter- 
fering frequency.  In  either  case,  each  trap  is 
effective  ovet  but  a small  frequency  range 
and  must  be  readjusted  for  other  frequencies. 

The  wave-trap  must  be  installed  as  close 
to  the  receiver  antenna  terminal  as  practic- 
able, hence  it  should  be  as  small  in  size  as 
possible.  The  variable  capacitor  may  be  a 
midget  air-tuned  trimmer  type,  and  the  coil 
may  be  wound  on  a 1-inch  dia.  form.  The  table 
of  figure  14  gives  winding  data  for  wave-traps 
built  around  standard  variable  capacitors.  For 
best  results,  both  a shunt  and  a series  trap 
should  be  employed  as  shown. 

Figure  15  shows  a two-circuit  coupled 
wave-trap  that  is  somewhat  sharper  in  tuning 
and  more  efficacious.  The  specifications  for 
the  secondary  coil  L,  may  be  obtained  from 
the  table  of  figure  14.  The  primary  coil  of  the 
shunt  trap  consists  of  3 to  5 closewound  turns 
of  the  same  size  wire  wound  in  the  same  di- 
rection on  the  same  form  as  Lj  and  separated 
from  the  latter  by  % of  an  inch. 

Overmodulation  A carrier  modulated  in  excess 
of  100  per  cent  acquires 
sharp  cutoff  periods  which  give  rise  to  tran- 
sients. These  transients  create  a broad  signal 
and  generate  spurious  responses.  Transients 
caused  by  overmodulation  of  a radio-telephone 
signal  may  at  the  same  time  bring  about  im- 
pact or  shock  excitation  of  nearby  receiving 
antennas  and  power  lines,  generating  inter- 
fering signals  in  that  manner. 


Broadcast  interference  due  to  overmodula- 
tion is  frequently  encountered.  The  remedy  is 
to  reduce  the  modulation  percentage  or  to  use 
a clipper-filter  system  or  a high-level  splatter 
suppressor  in  the  speech  circuit  of  the  trans- 
mitter. 

Cross  Cross  modulation  or  cross  talk  is 

Modulation  characterized  by  the  amateur  sig- 
nal riding  in  on  top  of  a strong 
broadcast  signal.  There  is  usually  no  hetero- 
dyne note,  the  amateur  signal  being  tuned  in 
and  out  with  the  program  carriers. 

This  effect  is  due  frequently  to  a faulty  in- 
put stage  in  the  affected  receiver.  Modulation 
of  the  interfering  carrier  will  swing  the  oper- 
ating point  of  the  input  tube.  This  type  of 
trouble  is  seldom  experienced  when  a varia- 
ble-g  tube  is  used  in  the  input  stage. 

Where  the  receiver  is  too  ancient  to  incor- 
porate such  a tube,  and  is  probably  poorly 
shielded  at  the  same  time,  it  will  be  better  to 
attach  a wave-trap  of  the  type  shown  in  figure 
12  rather  than  to  attempt  rebuilding  of  the  re- 
ceiver. The  addition  of  a good  ground  and  a 
shield  can  over  the  input  tube  often  adds  to 
the  effectiveness  of  the  wave-trap. 

Transmission  via  A small  amount  of  ca- 
Capacitive  Coupling  pacitive  coupling  is  now 
widely  used  in  receiver 
r.f.  and  antenna  transformers  as  a gain  booster 
at  the  high-frequency  end  of  the  tuning  range. 
The  coupling  capacitance  is  obtained  by 
means  of  a small  loop  of  wire  cemented  close 
to  the  grid  end  of  the  secondary  winding,  with 
one  end  directly  connected  to  the  plate  or  an- 
tenna end  of  the  primary  winding.  (See  figure 
16.) 


38  2 TV  and  Broadcast  Interference 


THE  RADIO 


CAPACITIVE 


Figure  16 

CAPACITIVE  BOOST  COUPLING 
CIRCUIT 

Such  circuits,  irtctuded  within  the  broadcast 
receiver  to  bring  up  the  stage  gain  at  the 
high-frequency  end  of  the  tuning  range,  have 
a tendency  to  increase  the  susceptibility  of 
the  receiver  to  interference  from  amateur- 
band  transmissions. 


It  is  easily  seen  that  a small  capacitor  at 
this  position  will  favor  the  coupling  of  the 
higher  frequencies.  This  type  of  capacitive 
coupling  in  the  receiver  coils  will  tend  to 
pass  amateur  high-frequency  signals  into  a re- 
ceiver tuned  to  broadcast  frequencies. 

The  amount  of  capacitive  coupling  may  be 
reduced  to  eliminate  interference  by  moving 
the  coupling  turn  further  away  from  the  sec- 
ondary coil.  However,  a simple  wave-trap  of 
the  type  shown  in  figure  12,  inserted  at  the 
antenna  input  terminal,  will  generally  accom- 
plish the  same  result  and  is  more  to  be  recom- 
mended than  reducing  the  amount  of  capaci- 
tive coupling  (which  lowers  the  receiver  gain 
at  the  high-frequency  end  of  the  broadcast 
band).  Should  the  wave-trap  alone  not  suffice, 
it  will  be  necessary  to  resort  to  a reduction 
in  the  coupling  capacitance. 

In  some  simple  broadcast  receivers,  capaci- 
tive coupling  is  obtained  by  closely  coupled 
primary  and  secondary  coils,  or  as  a result  of 
running  a long  primary  or  antenna  lead  close 
to  the  secondary  coil  of  an  unshielded  anten- 
na coupler. 

Phantoms  With  two  strong  local  carriers  ap- 
plied to  a non-linear  impedance, 
the  beat  note  resulting  from  cross-modulation 
between  them  may  fall  on  some  frequency 
within  the  broadcast  band  and  will  be  audible 
at  that  point.  If  such  a "phantom”  signal  falls 
on  a local  broadcast  frequency,  there  will  be 
heterodyne  interference  as  well.  This  is  a 
common  occurence  with  broadcast  receivers  in 
the  neighborhood  of  two  amateur  stations,  or 
an  amateur  and  a police  station.  It  also  some- 
times occurs  when  only  one  of  the  stations  is 
located  in  the  immediate  vicinity. 


As  an  example;  an  amateur  signal  on  3314 
kc.  might  beat  with  a local  2414-kc.  police 
carrier  to  produce  a 1100-kc.  phantom.  If  the 
two  carriers  are  strong  enough  in  the  vicinity 
of  a circuit  which  can  cause  rectification,  the 
1100-kc.  phantom  will  be  heard  in  the  broad- 
cast band.  A poor  contact  between  two  oxi- 
dized wires  can  produce  rectification. 

Two  stations  must  be  transmitting  simulta- 
neously to  produce  a phantom  signal;  when 
either  station  goes  off  the  air  tbe  phantom 
disappears.  Hence,  this  type  of  interference 
is  apt  to  be  reported  as  highly  intermittent  and 
might  be  difficult  to  duplicate  unless  a test 
oscillator  is  used  "on  location”  to  simulate 
the  missing  station.  Such  interference  cannot 
be  remedied  at  the  transmitter,  and  often  the 
rectification  takes  place  some  distance  from 
the  receivers.  In  such  occurrences  it  is  most 
difficult  to  locate  the  source  of  the  trouble. 

It  will  also  be  apparent  that  a phantom 
might  fall  on  the  intermediate  frequency  of  a 
simple  superhet  receiver  and  cause  interfer- 
ence of  the  untunable  variety  if  the  manufac- 
turer has  not  provided  an  i-f  wave-trap  in  the 
antenna  circuit. 

This  particular  type  of  phantom  may,  in 
addition  to  causing  i-f  interference,  generate 
harmonics  which  may  be  tuned  in  and  out  with 
heterodyne  whistles  from  one  end  of  the  re- 
ceiver dial  to  the  other.  It  is  in  this  manner 
that  birdies  often  result  from  the  operation  of 
nearby  amateur  stations. 

When  one  component  of  a phantom  is  a 
steady,  unmodulated  carrier,  only  the  intelli- 
gence present  on  the  other  carrier  is  conveyed 
to  the  broadcast  receiver. 

Phantom  signals  almost  always  may  be 
identified  by  the  suddenness  with  which  they 
ate  interrupted,  signalizing  withdrawal  of  one 
party  to  the  union.  This  is  especially  baffling 
to  the  inexperienced  interference-locater,  who 
observes  that  the  interference  suddenly  disap- 
pears, even  though  his  own  transmitter  re- 
mains in  operation. 

If  the  mixing  or  rectification  is  taking  place 
in  the  receiver  itself,  a phantom  signal  may 
be  eliminated  by  removing  either  one  of  the 
contributing  signals  from  the  receiver  input 
circuit.  A wave-trap  of  the  type  shown  in  fig- 
ure 12,  tuned  to  either  signal,  will  do  the 
trick.  If  the  rectification  is  taking  place  out- 
side the  receiver,  the  wave-trap  should  be 
tuned  to  the  frequency  of  the  phantom,  instead 
of  to  one  of  its  components.  I-f  wave-traps 
may  be  built  around  a 2.5-millihenry  r-f  choke 
as  the  inductor,  and  a compression-type  mica 
padding  capacitor.  The  capacitor  should  have 
a capacitance  range  of  250-525  ggfd.  for  the 
175-  and  206-kc.  intermediate  frequencies; 
65—175  ggfd.  for  260-kc.  and  other  intermedi- 


HANDBOOK 


Auto  Rectification  383 


ates  lying  between  250-  and  400-kc;  and 
17—80  ^gfd.  for  456-,  465-,  495-,  and  500-kc. 
Slightly  more  capacitance  will  be  requited  for 
resonance  with  a 2.1  millihenry  choke. 

Spurious  This  sort  of  interference  arises 
Emissions  from  the  transmitter  itself.  The 
radiation  of  any  signal  (other  than 
the  intended  carrier  frequency)  by  an  amateur 
station  is  prohibited  by  FCC  regulations.  Spu- 
rious radiation  may  be  traced  to  imperfect  neu- 
tralization, parasitic  oscillations  in  the  r-f  or 
modulator  stages,  or  to  "broadcast-band” 
variable-frequency  oscillators  or  e.c.o.’s. 

Low-frequency  parasitics  may  actually  oc- 
cur on  broadcast  frequencies  or  their  near  sub- 
harmonics, causing  direct  interference  to  pro- 
grams. An  all-wave  monitor  operated  in  the 
vicinity  of  the  transmitter  will  detect  these 
spurious  signals. 

The  remedy  will  be  obvious  in  individual 
cases.  Elsewhere  in  this  book  ate  discussed 
methods  of  complete  neutralization  and  the 
suppression  of  parasitic  oscillations  in  r-f 
and  audio  stages. 

A-c/d-c  Receivers  Ineitpensive  table-model 
a-c/d-c  receivers  are  par- 
ticularly susceptible  to  interference  from  ama- 
teur transmissions.  In  fact,  it  may  be  said 
with  a fait  degree  of  assurance  that  the  major- 
ity of  BCI  encountered  by  amateurs  operating 
in  the  1.8-Mc.  to  29-Mc.  range  is  a result  of 
these  inexpensive  receivers.  In  most  cases 
the  receivers  are  at  fault;  but  this  does  not 
absolve  the  amateur  of  his  responsibility  in 
attempting  to  eliminate  the  interference. 

Stray  Receiver  In  most  cases  of  interference 
Rectification  to  inexpensive  receivers,  par- 
ticularly those  of  the  a-c/d-c 
type,  it  will  be  found  that  stray  receiver  recti- 
fication is  causing  the  trouble.  The  offending 
stage  usually  will  be  found  to  be  a high-mu 
triode  as  the  first  audio  stage  following  the 
second  detector.  Tubes  of  this  type  are  quite 
non-linear  in  their  grid  characteristic,  and 
hence  will  readily  rectify  any  r-f  signal  ap- 
pearing between  grid  and  cathode.  The  r-f  sig- 
nal may  get  to  the  tube  as  a result  of  direct 
signal  pickup  due  to  the  lack  of  shielding,  but 
more  commonly  will  be  fed  to  the  tube  from 
the  power  line  as  a result  of  the  series  heater 
string. 

The  remedy  for  this  condition  is  simply  to 
insure  that  the  cathode  and  grid  of  the  high-mu 
audio  tube  (usually  a 12SQ7  oi  equivalent)  are 
at  the  same  r-f  potential.  This  is  accomplished 
by  placing  an  r-f  by-pass  capacitor  with  the 
shortest  possible  leads  directly  from  grid  to 
cathode,  and  then  adding  an  impedance  in  the 
lead  from  the  volume  control  to  the  grid  of  the 


HICH-MU  TUBE 


Figure  17 

CIRCUITS  FOR  ELIMINATING  AUDIO- 
STAGE RECTIFICATION 


audio  tube.  The  impedance  may  be  an  amateur 
band  r-f  choke  (such  as  a National  R-IOOU) 
for  best  results,  but  for  a majority  of  cases 
it  will  be  found  that  a 47,000-ohm  H-watt  re- 
sistor in  series  with  this  lead  will  give  satis- 
factory operation.  Suitable  circuits  for  such 
an  operation  on  the  receiver  are  given  in  fig- 
ure 17. 

In  many  a.c.-d.c.  receivers  there  is  no  r-f 
by-pass  included  across  the  plate  supply  recti- 
fier for  the  set.  If  there  is  an  appreciable 
level  of  r-f  signal  on  the  power  line  feeding 
the  receiver,  r-f  rectification  in  the  power 
rectifier  of  the  receiver  can  cause  a particu- 
larly bad  type  of  interference  which  may  be 
received  on  other  broadcast  receivers  in  the 
vicinity  in  addition  to  the  one  causing  the 
rectification.  The  soldering  of  a 0.01-pfd.  disc 
ceramic  capacitor  directly  from  anode  to  cath- 
ode of  the  power  rectifier  (whether  it  is  of  the 
vacuum-tube  or  selenium-rectifier  type)  usual- 
ly will  by-pass  the  r-f  signal  across  the  recti- 
fier and  thus  eliminate  the  difficulty. 

"Flooting"  Volume  Several  sets  have  been 

Control  Shafts  encountered  where  there 

was  only  a slightly  inter- 
fering signal;  but,  upon  placing  one’s  hand  up 
to  the  volume  control,  the  signal  would  great- 
ly increase.  Investigation  revealed  that  the 
volume  control  was  installed  with  its  shaft 
insulated  from  ground.  The  control  itself  was 
connected  to  a critical  part  of  a circuit,  in 
many  instances  to  the  grid  of  a high-gain  au- 
dio Stage.  The  cute  is  to  install  a volume  con- 
trol with  all  the  terminals  insulated  from  the 
shaft,  and  then  to  ground  the  shaft. 


384  TV  and  Broadcast  Interference 


THE  RADIO 


BAND 

COIL,  L 

CAPACITOR,  C 

3.5  Me. 

17  turns 
no.  14  enameled 
3>ineh  diameter 
2y4-inch  length 

1 0O-M/ifd. 

varioble 

7.0  Me. 

11  turns 
no.  14  enameled 
IVz’mch 
diameter 
IVi’inch  length 

lOO-^i/ifd. 

variable 

14  and 
21  Me. 

4 turns 

no.  10  enameled  100-;U/ifd. 
3>inch  diameter 

1 Vs-inch  length 

variable 

27  and 

28  Me. 

3 turns 
%-inch  o.d. 
eopper  tubing 
2-ineh  diameter 
1-ineh  length 

lOO'/iM^d. 

variable 

Figure  18 

COIL  AND  CAPACITOR  TABLE 
FOR  A-C  LINE  TRAPS 


Power-Line  When  radio-frequency  energy 

Pickup  from  a radio  transmitter  enters  a 

broadcast  receiver  through  the 
a-c  power  lines,  it  has  either  been  fed  back 
into  the  lighting  system  by  the  offending  trans- 
mitter, or  picked  up  from  the  air  by  over-head 
power  lines.  Underground  lines  are  seldom  re- 
sponsible for  spreading  this  interference. 

To  check  the  path  whereby  the  interfering 
signals  reach  the  line,  it  is  only  necessary  to 
replace  the  transmitting  antenna  with  a dum- 
my antenna  and  adjust  the  transmitter  for  max- 
imum output.  If  the  interference  then  ceases, 
overhead  lines  have  been  picking  up  the  en- 
ergy. The  trouble  can  be  cleared  up  by  install- 
ing a wave-trap  or  a commercial  line  filter  in 
the  power  lines  at  the  receiver.  If  the  receiver 
is  reasonably  close  to  the  transmitter,  it  is 
very  doubtful  that  changing  the  direction  of 
the  transmitting  antenna  to  right  angles  with 
the  overhead  lines  will  eliminate  the  trouble. 

If,  on  the  contrary,  the  interference  con- 
tinues when  the  transmitter  is  connected  to 
the  dummy  anterma,  radio-frequency  energy  is 
being  fed  directly  into  the  power  line  by  the 
transmitter,  and  the  station  must  be  inspected 
to  determine  the  cause. 

One  of  the  following  reasons  for  the  trouble 
will  usually  be  found:  ( 1)  the  t-f  stages  ate 
not  sufficiently  bypassed  and/or  choked,  (2) 
the  antenna  coupling  system  is  not  performing 
efficiently,  (3)  the  power  transformers  have 
no  electrostatic  shields;  or,  if  shields  are  pre- 
sent, they  are  ungrounded,  (4)  power  lines  are 
running  too  close  to  an  antenna  or  r-f  circuits 
carrying  high  currents.  If  none  of  these  causes 


METAL  BOX 


Figure  19 

RESONANT  POWER-LINE 
WAVE-TRAP  CIRCUIT 
The  resonant  type  of  power-line  filter  is 
more  effective  than  the  more  conventional 
**brute  force**  type  of  line  filter,  but  requires 
tuning  to  the  operating  frequency  of  the 
transmitter. 


af^ly,  wave-traps  must  be  installed  in  the 
power  lines  at  the  transmitter  to  remove  r-f 
energy  passing  back  into  the  lighting  system. 

The  wave-traps  used  in  the  power  lines  at 
transmitter  or  receiver  must  be  capable  of 
passing  relatively  high  current.  The  coils  are 
accordingly  wound  with  heavy  wire.  Figure  18 
lists  the  specifications  for  power  line  wave- 
trap  coils,  while  figure  19  illustrates  the  meth- 
od of  connecting  these  wave-traps.  Observe 
that  these  traps  are  enclosed  in  a shield  box 
of  heavy  iron  or  steel,  well  grounded. 

All-Wove  Each  complete-coverage  home  re- 
Receivers  ceiver  is  a potential  source  of  an- 
noyance to  the  transmitting  ama- 
teur. The  novice  short-wave  broadcast-listener 
who  tunes  in  an  amateur  station  often  con- 
siders it  an  interfering  signal,  and  complains 
accordingly. 

Neither  selectivity  nor  image  rejection  in 
most  of  these  sets  is  comparable  to  those 
properties  in  a communication  receiver.  The 
result  is  that  an  amateur  signal  will  occupy 
too  much  dial  space  and  appear  at  more  than 
one  point,  giving  rise  to  interference  on  ad- 
jacent channels  and  distant  channels  as  well. 

If  carrier-frequency  harmonics  are  present 
in  the  amateur  transmission,  serious  interfer- 
ence will  result  at  the  all-wave  receiver.  The 
harmonics  may,  if  the  carrier  frequency  has 
been  so  unfortunately  chosen,  fall  directly 
upon  a favorite  short-wave  broadcast  station 
and  arouse  warranted  objection. 

The  amateur  is  apt  to  be  blamed,  too,  for 
transmissions  for  which  he  is  not  responsible, 
so  great  is  the  public  ignorance  of  short-wave 
allocations  and  signals.  Owners  of  all-wave 
receivers  have  been  quick  to  ascribe  to  ama- 
teur stations  all  signals  they  heat  from  tape 
machines  and  V-wheels,  as  well  as  stray  tones 
and  heterodyne  flutters. 


HANDBOOK 


Image  Interference 


385 


The  amateur  cannot  be  held  responsible 
when  his  carrier  is  deliberately  tuned  in  on 
an  all-wave  receiver.  Neither  is  he  account- 
able for  the  width  of  his  signal  on  the  receiver 
dial,  or  for  the  strength  of  image  repeat  points, 
if  it  can  be  proven  that  the  receiver  design 
does  not  afford  good  selectivity  and  image  re- 
jection. 

If  he  so  desires,  the  amateur  (or  the  owner 
of  the  receiver)  might  sharpen  up  the  received 
signal  somewhat  by  shortening  the  receiving 
antenna.  Set  retailers  often  supply  quite  a 
sizeable  antenna  with  all- wave  receivers,  but 
most  of  the  time  these  sets  perform  almost  as 
well  with  a few  feet  of  inside  antenna. 

The  amateur  is  accountable  for  harmonics 
of  his  carrier  frequency.  Such  emissions  are 
unlawful  in  the  first  place,  and  he  must  take 
all  steps  necessary  to  their  suppression.  Prac- 
tical suggestions  for  the  elimination  of  har- 
monics have  been  given  earlier  in  this  chap- 
ter under  Television  Interference. 


Image  Interference  In  addition  to  those  types 
of  interference  already 
discussed,  there  are  two  more  which  are  com- 
mon to  superhet  receivers.  The  prevalence  of 
these  types  is  of  great  concern  to  the  ama- 
teur, although  the  responsibility  for  their  ex- 
istence more  properly  rests  with  the  broadcast 
receiver. 

The  mechanism  whereby  image  production 
takes  place  may  be  explained  in  the  following 
manner:  when  the  first  detector  is  set  to  the 
frequency  of  an  incoming  signal,  the  high-fre- 
quency oscillator  is  operating  on  another  fre- 
quency which  differs  from  the  signal  by  the 
number  of  kilocycles  of  the  intermediate  fre- 
quency. Now,  with  the  setting  of  these  two 
stages  undisturbed,  there  is  another  signal 
which  will  beat  with  the  high-frequency  oscil- 
lator to  produce  an  i-f  signal.  This  other  sig- 
nal is  the  so-called  image,  which  is  separated 
from  the  desired  signal  by  twice  the  inter- 
mediate frequency. 

Thus,  in  a receiver  with  175-kc.  i.f.,  tuned 
to  1000  kc.:  the  h-f  oscillator  is  operating  on 
1175  kc. , and  a signal  on  1350  kc.  (1000  kc. 
plus  2 X 175  kc.)  will  beat  with  this  1175  kc. 
oscillator  frequency  to  produce  the  175-kc.  i-f 
signal.  Similarly,  when  the  same  receiver  is 
tuned  to  1400  kc.,  an  amateur  signal  on  1750 
kc.  can  come  through. 

If  the  image  appears  only  a few  cycles  or 
kilocycles  from  a broadcast  carrier,  heterodyne 
interference  will  be  present  as  well.  Other- 
wise, it  will  be  tuned  in  and  out  in  the  manner 
of  a station  operating  in  the  broadcast  band. 
Sharpness  of  tuning  will  be  comparable  to  that 
of  broadcast  stations  producing  the  same  a-v-c 
voltage  at  the  receiver. 


The  second  variety  of  superhet  interference 
is  the  result  of  harmonics  of  the  receiver  h-f 
oscillator  beating  with  amateur  carriers  to  pro- 
duce the  intermediate  frequency  of  the  receiv- 
er. The  amateur  transmitter  will  always  be 
found  to  be  on  a frequency  equal  to  some  har- 
monic of  the  receiver  h-f  oscillator,  plus  or 
minus  the  intermediate  frequency. 

As  an  example:  when  a broadcast  superhet 
with  465-kc.  i.f.  is  tuned  to  1000  kc.,  its  high- 
frequency  oscillator  operates  on  1465  kc.  The 
third  harmonic  of  this  oscillator  frequency  is 
4395  kc.,  which  will  beat  with  an  amateur  sig- 
nal on  3930  kc.  to  send  a signal  through  the 
i-f  amplifier.  The  3930  kc.  signal  would  be 
tuned  in  at  the  1000-kc.  point  on  the  dial. 

Some  oscillator  harmonics  are  so  related  to 
amateur  frequencies  that  more  than  one  point 
of  interference  will  occur  on  the  receiver  dial. 
Thus,  a 3500-kc.  signal  may  be  tuned  in  at 
six  points  on  the  dial  of  a nearby  broadcast 
superhet  having  175  kc.  i.f.  and  no  r-f  stage. 

Insofar  as  remedies  for  image  and  harmonic 
superhet  interference  are  concerned,  it  is  well 
to  remember  that  if  the  amateur  signal  did  not 
in  the  first  place  reach  the  input  stage  of  the 
receiver,  the  annoyance  would  not  have  been 
created.  It  is  therefore  good  policy  to  try  to 
eliminate  it  by  means  of  a wave-trap  or  low- 
pass  filter.  Broadcast  superhets  are  not  al- 
ways the  acme  of  good  shielding,  however, 
and  the  amateur  signal  is  apt  to  enter  the  cir- 
cuit through  channels  other  than  the  input  cir- 
cuit. If  a wave-trap  or  filter  will  not  cure  the 
trouble,  the  only  alternative  will  be  to  attempt 


-pm o 

i i OUTPU' 

— 1 1 O 


CONSTANT  K TYPE 


Figure  20 

TYPES  OF  LOW-PASS  FILTERS 

Pilfers  such  as  these  may  be  used  in  the 
circuit  between  the  antenna  and  the  input 
of  the  receiver. 


386  TV  and  Broadcast  Interference 


ANT.  L1  L2 


GND. 

Figure  21 

COMPOSITE  LOW-PASS  FILTER 
CIRCUIT 

This  filter  is  highly  effective  in  reducing 
broadcast  interference  from  all  high  frequen- 
cy  stations,  and  requires  no  tuning.  Con- 
stants for  400  ohm  terminal  impedance  and 
1600  kc.  cutoff  are  as  follows:  Lj,  65  turns 
no.  22  d.c.c.  closewound  on  1 Vi  in.  dia.  form. 
L2,  41  turns  ditto,  not  coupled  to  Lj.  Cj, 
250  m^fd.  fixed  mica  capacitor,  C2,  400  (m-fd. 
fixed  mica  copacitor.  Cj  anci  C4,  150  pp.fd. 
fixed  mica  capacitors,  former  of  5%  toler- 
ance. With  some  receivers,  better  results 
will  Se  obtained  with  a 200  ohm  carbon  re- 
sistor inserted  between  the  filter  and  an- 
tenna post  on  the  receiver.  With  other  re- 
ceivers the  effectiveness  will  be  improved 
with  a 600  ohm  carbon  resistor  placed  from 
the  antenna  post  to  the  ground  post  on  the 
receiver.  The  filter  should  be  placed  as 
close  to  the  receiver  terminals  as  possible. 


to  select  a transmitter  frequency  such  that 
neither  image  nor  harmonic  interference  will  be 
set  up  on  favorite  stations  in  the  susceptible 
receivers.  The  equation  given  earlier  may  be 
used  to  determine  the  proper  frequencies. 

Low  Pass  Filters  The  greatest  drawback  of 
the  wave-trap  is  the  fact 
that  it  is  a single-frequency  device;  i.e.— it 
may  be  set  to  reject  at  one  time  only  one  fre- 
quency (or,  at  best,  an  extremely  narrow  band 
of  frequencies).  Each  time  the  frequency  of 
the  interfering  transmitter  is  changed,  every 
wave-trap  tuned  to  it  must  be  retuned.  A much 
more  satisfactory  device  is  the  wave  filter 
which  requires  no  tuning.  One  type,  the  low- 
pass  filter,  passes  all  frequencies  below  one 
critical  frequency,  and  eliminates  all  higher 
frequencies.  It  is  this  property  that  makes  the 
device  ideal  for  the  task  of  removing  amateur 
frequencies  from  broadcast  receivers. 

A good  low-pass  filter  designed  for  maxi- 
mum attenuation  around  1700  kc.  will  pass 
all  broadcast  carriers,  but  will  reject  signals 
originating  in  any  amateur  band.  Naturally 
such  a device  should  be  installed  only  in 
standard  broadcast  receivers,  never  in  all- 
wave sets. 


Two  types  of  low-pass  filter  sections  are 
shown  in  figure  20.  A composite  arrangement 
comprising  a section  of  each  type  is  more 
effective  than  either  type  operating  alone.  A 
composite  filter  composed  of  one  K-section 
and  one  shunt-derived  M-section  is  shown  in 
figure  21,  and  is  highly  recommended.  The 
M-section  is  designed  to  have  maximum  atten- 
uation at  1700  kc.,  and  for  that  reason  C3 
should  be  of  the  **close  tolerance**  variety. 
Likewise,  C3  should  not  be  stuffed  down  in- 
side Lj  in  the  interest  of  compactness,  as 
this  will  alter  the  inductance  of  the  coil  appre- 
ciably, and  likewise  the  resonant  frequency. 

If  a fixed  150  ^^fd.  mica  capacitor  of  5 per 
cent  tolerance  is  not  available  for  Cj,  a com- 
pression trimmer  covering  the  range  of  125— 
175  may  be  substituted  and  adjusted  to 

give  maximum  attenuation  at  about  1700  kc. 

19-5  HI-FI  Interference 

The  rapid  growth  of  high-fidelity  sound  systems 
in  the  home  has  brought  about  many  cases  of 
interference  from  a nearby  amateur  transmitter. 
In  most  cases,  the  interference  is  caused  by 
stray  pickup  of  the  r-f  signal  by  the  interconnect- 
ing leads  of  the  hi-fi  system  and  audio  rectifi- 
cation in  the  low  level  stages  of  the  amplifier. 
The  solution  to  this  difficulty,  in  general,  is  to 
bypass  and  filter  all  speaker  and  power  leads 
to  the  hi-fi  amplifier  and  preamplifier.  A com- 
bination of  a VHF  choke  and  500  fifild  ceramic 
disc  capacitors  in  each  power  and  speaker  lead 
will  eliminate  r-f  pickup  in  the  high  level  section 
of  the  amplifier.  A filter  such  as  shown  in  figure 
17A  placed  in  the  input  circuit  of  the  first  audio 
stage  of  the  preamplifier  will  reduce  the  level  of 
the  r-f  signal  reaching  the  input  circuit  of  the 
amplifier.  To  prevent  loss  of  the  higher  audio 
frequencies  it  may  be  necessary  to  decrease  the 
value  of  the  grid  bypass  capacitor  to  50  nfiid 
or  so. 

Shielded  leads  should  be  employed  between  the 
amplifier  and  the  turntable  or  f-m  tuner.  The 
shield  should  be.  grounded  at  both  ends  of  the 
line  to  the  chassis  of  the  equipment,  and  cate 
should  be  taken  to  see  that  the  line  does  not 
approach  an  electrical  half-wave  length  of  the 
radio  signal  causing  the  interference.  In  some 
instances,  shielding  the  power  cable  to  the  hi-fi 
equipment  will  aid  in  reducing  interference.  The 
framework  of  the  phonograph  turntable  should  be 
grounded  to  the  chassis  of  the  amplifier  to  reduce 
stray  r-f  pickup  in  the  turntable  equipment. 


CHAPTER  TWENTY 


Transmitter  Keying  and  Control 


20-1  Power  Systems 

It  is  probable  that  the  average  amateur  sta- 
tion that  has  been  in  operation  for  a number 
of  years  will  have  at  least  two  transmitters 
available  for  operation  on  different  frequency 
bands,  at  least  two  receivers  or  one  receiver 
and  a converter,  at  least  one  item  of  monitor- 
ing or  frequency  measuring  equipment  and 
probably  two,  a v.f.o.,  a speech  amplifier,  a 
desk  light,  and  a clock.  In  addition  to  the 
above  8 or  10  items,  there  must  be  an  outlet 
available  for  a soldering  iron  and  there  should 
be  one  or  two  additional  outlets  available  for 
plugging  in  one  or  two  pieces  of  equipment 
which  ate  being  worked  upon. 

It  thus  becomes  obvious  that  10  or  12  out- 
lets connected  to  the  115-volt  a-c  line  should 
be  available  at  the  operating  desk.  It  may  be 
practicable  to  have  this  number  of  outlets  in- 
stalled as  an  outlet  strip  along  the  baseboard 
at  the  time  a new  home  is  being  planned  and 
constructed.  Or  it  might  be  well  to  install 
the  outlet  strip  on  the  operating  desk  so  as 
to  have  the  flexibility  of  moving  the  operating 
desk  from  one  position  to  another.  Alterna- 
tively, the  outlet  strip  might  be  wall  mounted 
just  below  the  desk  top. 

Power  Droin  When  the  power  drain  of  all  the 
Per  Outlet  items  of  equipment,  other  than 

transmitters,  used  at  the  oper- 
ating position  is  totalled,  you  probably  will 
find  that  350  to  600  watts  will  be  required. 


Since  the  usual  home  outlet  is  designed  to 
handle  only  about  600  watts  maximum,  the 
transmitter,  unless  it  is  of  relatively  low  power, 
should  be  powered  from  another  source.  This 
procedure  is  desirable  in  any  event  so  that  the 
voltage  supplied  to  the  receiver,  frequency  con- 
trol, and  frequency  monitor  will  be  substan- 
tially constant  with  the  transmitter  on  or  off 
the  air. 

So  we  come  to  two  general  alternative  plans 
with  their  variations.  Plan  (A)  is  the  more  de- 
sirable and  also  the  most  expensive  since  it 
involves  the  installation  of  two  separate  lines 
from  the  meter  box  to  the  operating  position 
either  when  the  house  is  constructed  or  as  an 
alteration.  One  line,  with  its  switch,  is  for  the 
transmitters  and  the  other  line  and  switch  is 
for  receivers  and  auxiliary  equipment.  Plan 
( B)is  the  more  practicable  for  the  average  ama- 
teur, but  its  use  requires  that  all  cords  be  re- 
moved from  the  outlets  whenever  the  station 
is  not  in  use  in  order  to  comply  with  the  elec- 
trical codes. 

Figure  1 shows  a suggested  arrangement  for 
carrying  out  Plan  (A).  In  most  cases  an  instal- 
lation such  as  this  will  require  approval  of  the 
plans  by  the  city  or  county  electrical  inspector. 
Then  the  installation  itself  will  also  requite 
inspection  after  it  has  been  completed.  It  will 
be  necessary  to  use  approved  outlet  boxes  at 
the  tear  of  the  transmitter  where  the  cable  is 
connected,  and  also  at  the  operating  bench 

where  the  other  BX  cable  connects  to  the  out- 
let strip.  Also,  the  connectors  at  the  rear  of 
the  transmitter  will  have  to  be  of  an  approved 


387 


388  Transmitter  Keying  and  Control 


THE  RADIO 


THE  PLAN  (A)  POWER  SYSTEM 


A-c  line  power  from  the  main  fuse  box  in  the 
house  is  run  separately  to  the  receiving 
equipment  and  to  the  transmitting  equipment. 
Separate  switches  and  fuse  blocks  then  are 
available  for  the  transmitters  and  for  the 
auxiliary  equipment.  5/nce  the  fuses  In  the 
boxes  at  the  operating  room  will  be  In  series 
with  those  at  the  main  fuse  box,  those  in  the 
operating  room  should  have  a lower  rating 
than  those  at  the  main  fuse  box.  Then  it  will 
always  be  possible  to  rep/ace  blown  fuses 
without  leaving  the  operating  room.  The  fuse 
boxes  can  conveniently  be  located  alongside 
one  another  on  the  wall  of  the  operating  room. 


PLAN  (b) 


Figure  2 

THE  PLAN  (B)  POWER  SYSTEM 
This  system  is  less  convenient  than  the  (A) 
system,  but  does  not  require  extensive  re- 
wiring of  the  electrical  system  within  the 
house  to  accommodate  the  arrangement.  Thus 
it  is  better  for  a temporary  or  semi-permanent 
installation.  In  most  cases  it  will  be  neces- 
sary to  run  an  extra  conduit  from  the  main 
fuse  box  to  the  outlet  from  which  the  trans- 
mitter is  powered,  since  the  standard  arrange- 
ment in  most  houses  is  to  run  all  the  outlets 
in  one  room  (and  sometimes  all  In  the  house) 
from  a single  pair  of  fuses  and  leads. 


type.  It  is  possible  also  that  the  BX  cable  will 
have  to  be  permanently  affixed  to  the  trans- 
mitter with  the  connector  at  the  fuse-box  end. 
These  details  may  be  worked  out  in  advance 
with  the  electrical  inspector  for  your  area. 

The  general  aspects  of  Plan  ( B)  are  shown 
in  figure  2.  The  basic  difference  between  the 
two  plans  is  that  (A)  represents  a permanent 
installation  even  though  a degree  of  mobility 
is  allowed  through  the  use  of  BX  for  power 
leads,  while  plan  (B)  is  definitely  a temporary 
type  of  installation  as  far  as  the  electrical  in- 
spector is  concerned.  While  it  will  be  permis- 
sible in  most  areas  to  leave  the  transmitter 
cord  plugged  into  the  outlet  even  though  it  is 
turned  off,  the  Fire  Insurance  Underwriters 
codes  will  make  it  necessary  that  the  cord 
which  runs  to  the  group  of  outlets  at  the  back 
of  the  operating  desk  be  removed  whenever  the 
equipment  is  not  actually  in  use. 

Whether  the  general  aspects  of  plans  (A)  or 
(B)  are  used  it  will  be  necessary  to  run  a num- 
ber of  control  wires,  keying  and  audio  leads, 
and  an  excitation  cable  from  the  operating  desk 


to  the  transmitter.  Control  and  keying  wires 
can  best  be  grouped  into  a multiple-wire  rubber- 
covered  cable  between  the  desk  and  the  trans- 
mitter. Such  an  arrangement  gives  a good  ap- 
pearance, and  is  particularly  practical  if  cable 
connectors  are  used  at  each  end.  High-level 
audio  at  a moderate  impedance  level  (600  ohms 
or  below)  may  be  run  in  the  same  control  cable 
as  the  other  leads.  However,  low-level  audio 
can  best  be  run  in  a small  coaxial  cable.  Small 
coaxial  cable  such  as  RG-58/U  or  RG-59/U 
also  is  quite  satisfactory  and  quite  convenient 
for  the  signal  from  the  v.f.o.  to  the  r-f  stages 
in  the  transmitter.  Coaxial-cable  connectors  of 
the  UG  series  are  quite  satisfactory  for  the 
terminations  both  for  the  v-f-o  lead  and  for  any 
low-level  audio  cables. 

Checking  an  To  make  sure  that  an  outlet  will 
Outlet  with  a stand  the  full  load  of  the  entire 
Heavy  Load  transmitter,  plug  in  an  electric 
heater  rated  at  about  50  per  cent 
greater  wattage  than  the  power  you  expect  to 
draw  from  the  line.  If  the  line  voltage  does 


HANDBOOK 


Transmitter  Control  389 


not  drop  more  than  5 volts  (assuming  a 117- 
volt  line)  under  load  and  the  wiring  does  not 
overheat,  the  wiring  is  adequate  to  supply  the 
transmitter.  About  600  watts  total  drain  is  the 
maximum  that  should  be  drawn  from  a 117-volt 
lighting  outlet  or  circuit.  For  greater  power, 
a separate  pair  of  heavy  conductors  should  be 
run  right  from  the  meter  box.  For  a 1-kw.  phone 
transmitter  the  total  drain  is  so  great  that  a 
230-volt  "split”  system  ordinarily  will  be  re- 
quired. Most  of  the  newer  homes  ate  wired  with 
this  system,  as  are  homes  utilizing  electricity 
for  cooking  and  heating. 

With  a three-wire  system,  be  sure  there  is 
no  fuse  in  the  neutral  wire  at  the  fuse  box.  A 
neutral  fuse  is  not  required  if  both  "hot”  legs 
are  fused,  and,  should  a neutral  fuse  blow, 
there  is  a chance  that  damage  to  the  radio 
transmitter  will  result. 

If  you  have  a high  power  transmitter  and  do 
a lot  of  operating,  it  is  a good  idea  to  check 
on  your  local  power  rates  if  you  are  on  a 
straight  lighting  rate.  In  some  cities  a lower 
rate  can  be  obtained  (but  with  a higher  "mini- 
mum”) if  electrical  equipment  such  as  an 
electric  heater  drawing  a specified  amount 
of  current  is  petmanently  wired  in.  It  is  not 
required  that  you  use  this  equipment,  merely 
that  it  be  permanently  wired  into  the  electrical 
system.  Naturally,  however,  there  would  be  no 
saving  unless  you  expect  to  occupy  the  same 
dwelling  for  a considerable  length  of  time. 

Outlet  Strips  The  outlet  strips  which  have 
been  suggested  for  installation 
in  the  baseboard  or  for  use  on  the  rear  of  a desk 
are  obtainable  from  the  large  electrical  supply 
houses.  If  such  a house  is  not  in  the  vicinity 
it  is  probable  that  a local  electrical  contractor 
can  order  a suitable  type  of  strip  from  one  of 
the  supply  house  catalogs.  These  strips  are 
quite  convenient  in  that  they  are  available  in 
varying  lengths  with  provision  for  inserting 
a-c  line  plugs  throughout  their  length.  The 
a-c  plugs  from  the  various  items  of  equipment 
on  the  operating  desk  then  may  be  inserted 
in  the  outlet  strip  throughout  its  length.  In 
many  cases  it  will  be  desirable  to  reduce  the 
equipment  cord  lengths  so  that  they  will  plug 
neatly  into  the  outlet  strip  without  an  excess 
to  dangle  behind  the  desk. 

Contactors  and  The  use  of  power-control  con- 
Relays  tactors  and  relays  often  will 

add  considerably  to  the  oper- 
ating convenience  of  the  station  installation. 
The  most  practicable  arrangement  usually  is 
to  have  a main  a-c  line  switch  on  the  front  of 
the  ttansmitter  to  apply  power  to  the  filament 
transformers  and  to  the  power  control  circuits. 
It  also  will  be  found  quite  convenient  to  have 
a single  a-c  line  switch  on  the  operating  desk 


to  energize  or  cut  the  power  from  the  outlet 
strip  on  the  teat  of  the  operating  desk.  Through 
the  use  of  such  a switch  it  is  not  necessary  to 
remember  to  switch  off  a large  number  of  sepa- 
rate switches  on  each  of  the  items  of  equip- 
ment on  the  operating  desk.  The  alternative 
arrangement,  and  that  which  is  approved  by  the 
Underwriters,  is  to  remove  the  plugs  from  the 
wall  both  for  the  transmitter  and  for  the  oper- 
ating-desk outlet  strip  when  a period  of  oper- 
ation has  been  completed. 

While  the  insertion  of  plugs  or  operation  of 
switches  usually  will  be  found  best  for  ap- 
plying the  a-c  line  power  to  the  equipment,  the 
changing  over  between  transmit  and  receive 
can  best  be  accomplished  through  the  use  of 
relays.  Such  a system  usually  involves  thtee 
relays,  or  three  groups  of  relays.  The  relays 
and  their  functions  are:  (1)  power  control  relay 
for  the  ttansmitter— applies  11 5- volt  line  to  the 
primary  of  the  high-voltage  transformer  and 
turns  on  the  exciter;  (2)  control  relay  for  the 
receiver— makes  the  receiver  inoperative  by 
any  one  of  a number  of  methods  when  closed, 
also  may  apply  power  to  the  v.f.o.  and  to  a 
keying  or  a phone  monitot;  and  (3)  the  antenna 
changeover  relay— connects  the  antenna  to  the 
transmitter  when  the  transmitter  is  energized 
and  to  the  receiver  when  the  transmitter  is  not 
operating.  Several  circuits  illustrating  the  ap- 
plication of  relays  to  such  control  arrangements 
are  discussed  in  the  paragraphs  to  follow  in 
this  chapter. 

Controlling  Transmitter  It  is  necessary,  in 
Power  Output  order  to  comply  with 

FCC  regulations,  that 
transmitter  power  output  be  limited  to  the  mini- 
mum amount  necessary  to  sustain  communica- 
tion. This  requirement  may  be  met  in  several 
ways.  Many  amateurs  have  two  transmitters; 
one  is  capable  of  relatively  high  power  output 
for  use  when  calling,  or  when  interference  is 
severe,  and  the  other  is  capable  of  consider- 
ably less  power  output.  In  many  cases  the 
lower  powered  transmitter  acts  as  the  exciter 
for  the  higher  powered  stage  when  full  power 
output  is  requited.  But  the  majority  of  the  ama- 
teurs using  a high  powered  equipment  have 
some  provision  for  reducing  the  plate  voltage 
on  the  high-level  stages  when  reduced  power 
output  is  desired. 

One  of  the  most  common  arrangements  for 
obtaining  two  levels  of  power  output  involves 
the  use  of  a plate  transformer  having  a double 
primary  for  the  high-voltage  power  supply.  The 
majority  of  the  high-power  plate  transformers 
of  standard  manufacture  have  just  such  a dual- 
primary  arrangement.  The  two  primaries  are 
designed  for  use  with  either  a 115-volt  or  230- 
volt  line.  When  such  a transformer  is  to  be 
operated  from  a 115-volt  line,  operation  of  both 


Transmitter  Keying  and  Control 


THE  RADIO 


POWER  CONTROL  SYSTEMS 
The  circuit  at  (A)  is  for  use  with  a I75-vo/t 
o-c  line.  Transformer  T is  of  the  standord 
type  having  two  llS-volt  primaries;  these 
primaries  are  connected  in  series  for  half- 
voltage  output  when  the  power  control  relay 
K;  is  energized  but  the  hi~lo  relay  K2  is  not 
operated.  When  both  relays  are  energized  the 
full  output  voltage  is  obtained.  At  (8)  is  a 
circuit  for  use  with  a standard  230.volt  resi- 
dence line  with  grounded  neutral.  The  two 
relays  control  the  output  of  the  power  sup- 
plies the  same  as  at  (A). 


primaries  in  parallel  will  deliver  full  output 
from  the  plate  supply.  Then  when  the  two  pri- 
maries are  connected  in  series  and  still  oper- 
ated from  the  115- volt  line  the  output  voltage 
from  the  supply  will  be  reduced  approximately 
to  one  half.  In  the  case  of  the  normal  class  C 
amplifier,  a reduction  in  plate  voltage  to  one 
half  will  reduce  the  power  input  to  the  stage 
to  one  quarter. 

If  the  transmitter  is  to  be  operated  from  a 
230-volt  line,  the  usual  procedure  is  to  operate 
the  filaments  from  one  side  of  the  line,  the 


low-voltage  power  supplies  from  the  other  side, 
anti  the  primaries  of  the  high-voltage  trans- 
former across  the  whole  line  for  full  power 
output.  Then  when  reduced  power  output  is 
required,  the  primary  of  the  high-voltage  plate 
transformer  is  operated  from  one  side  to  center 
tap  rather  than  across  the  whole  line.  This 
procedure  places  115  volts  across  the  230-voIt 
winding  the  same  as  in  the  case  discussed  in 
the  previous  paragraph.  Figure  3 illustrates 
the  two  standard  methods  of  power  reduction 
with  a plate  transformer  having  a double  pri- 
mary; (A)  shows  the  connections  for  use  with 
a 115-volt  line  and  (B)  shows  the  arrangement 
for  a 230-volt  a-c  power  line  to  the  transmitter. 

The  full-voltage/half-voltage  methods  for 
controlling  the  power  input  to  the  transmitter, 
as  just  discussed,  are  subject  to  the  limitation 
that  only  two  levels  of  power  input  (full  power 
and  quarter  power)  are  obtainable.  In  many 
cases  this  will  be  found  to  be  a limitation  to 
flexibility.  When  tuning  the  transmitter,  the 
antenna  coupling  network,  or  the  antenna  sys- 
tem itself  it  is  desirable  to  be  able  to  reduce 
the  power  input  to  the  final  stage  to  a rela- 
tively low  value.  And  it  is  further  convenient 
to  be  able  to  vary  the  power  input  continuous- 
ly from  this  relatively  low  input  up  to  the  full 
power  capabilities  of  the  transmitter.  The  use 
of  a variable-ratio  auto-transformer  in  the  cir- 
cuit from  the  line  to  the  primary  of  the  plate 
transformer  will  allow  a continuous  variation 
in  power  input  from  zero  to  the  full  capability 
of  the  transmitter. 

Variable-Ratio  There  are  several  types 

Auto-Transformers  of  variable-ratio  auto-trans- 
formers available  on  the 
market.  Of  these,  the  most  common  are  the 
Variac  manufactured  by  the  General  Radio 
Company,  and  the  Powerstat  manufactured  by 
the  Superior  Electric  Company.  Both  these 
types  of  variable-ratio  transformers  are  excel- 
lently constructed  and  are  available  in  a wide 
range  of  power  capabilities.  Each  is  capable 
of  controlling  the  line  voltage  from  zero  to 
about  15  per  cent  above  the  nominal  line  volt- 
age. Each  manufacturer  makes  a single-phase 
unit  capable  of  handling  an  output  power  of 
about  175  watts,  one  capable  of  about  750  to 
800  watts,  and  a unit  capable  of  about  1500  to 
1800  watts.  The  maximum  power-output  capa- 
bility of  these  units  is  available  only  at  ap- 
proximately the  nominal  line  voltage,  and  must 
be  reduced  to  a maximum  current  limitation 
when  the  output  voltage  is  somewhat  above  or 
below  the  input  line  voltage.  This,  however,  is 
not  an  important  limitation  for  this  type  of 
application  since  the  output  voltage  seldom 
will  be  raised  above  the  line  voltage,  and  when 
the  output  voltage  is  reduced  below  the  line 
voltage  the  input  to  the  transmitter  is  reduced 
accordingly. 


HANDBOOK 


Transmitter  Control  391 


TO  EXCITER  POWER  SUPPLIES 


TOH.V. 

POWER  SUPPLY 


115V.A.C.  ^ 

LINE 


Figure  4 

CIRCUIT  WITH  VARIABLE-RATIO 
AUTO-TRANSFORMER 
When  the  dummy  plug  is  inserted  into  the  re- 
ceptacle on  the  equipment,  closing  of  the 
power  control  relay  will  apply  full  voltage 
to  the  primaries.  With  the  cable  from  the 
Variac  or  Powerstat  plugged  into  the  socket 
the  voltage  output  of  the  high-voltage  power 
supply  may  be  varied  from  zero  to  about  IS 
per  cent  above  normal. 


TRANSMITTER 
> FILAMENT 
TRANSFORMERS 


1115  V.  TO  EXCITER  AND 
HIGH-VOLTAGE  RELAYS, 
AND  TO  RECEIVER  CON- 
TROL AND  ANTENNA 
CHANGEOVER  RELAYS 


PROTECTIVE  CONTROL  CIRCUIT 


With  this  circuit  orrongemenf  either  switch 
may  be  closed  first  to  light  the  heaters  of 
all  tubes  and  the  filament  pilot  light.  Then 
when  the  second  switch  is  closed  the  high 
voltage  will  be  applied  to  the  transmitter  and 
the  red  pilot  will  light.  With  o 30-second  de- 
lay between  the  closing  of  the  first  switch 
and  the  closing  of  the  second,  the  rectifier 
tubes  will  be  adequately  protected.  Similarly, 
the  opening  of  either  switch  will  remove 
plate  voltage  from  the  rectifiers  while  the 
heaters  remain  lighted. 


One  convenient  arrangement  for  using  a 
Variac  or  Powerstat  in  conjunction  with  the 
high-voltage  transformer  of  a transmitter  is 
illustrated  in  figure  4.  In  this  circuit  a heavy 
three-wire  cable  is  run  from  a plug  on  the  trans- 
mitter to  the  Variac  or  Powerstat.  The  Variac 
or  Powerstat  then  is  installed  so  that  it  is  ac- 
cessible from  the  operating  desk  so  that  the 
input  power  to  the  transmitter  may  be  con- 
trolled during  operation.  If  desired,  the  cable 
to  the  Variac  or  Powerstat  may  be  unplugged 
from  the  transmitter  and  a dummy  plug  inserted 
in  its  place.  With  the  dummy  plug  in  place  the 
transmitter  will  operate  at  normal  plate  voltage. 
This  arrangement  allows  the  transmitter  to  be 
wired  in  such  a manner  that  an  external  Variac 
or  Powerstat  may  be  used  if  desired,  even 
though  the  unit  is  not  available  at  the  time 
that  the  transmitter  is  constructed. 

Notes  on  the  Use  Plate  voltage  to  the  modula- 
of  the  Voriac  tors  may  be  controlled  at  the 

or  Powerstat  same  time  as  the  plate  volt- 

age to  the  final  amplifier  is 
varied  if  the  modulator  stage  uses  beam  tetrode 
tubes;  variation  in  the  plate  voltage  on  such 
tubes  used  as  modulators  causes  only  a mod- 
erate change  in  the  standing  plate  current. 
Since  the  final  amplifier  plate  voltage  is  being 
controlled  simultaneously  with  the  modulator 


plate  voltage,  the  conditions  of  impedance 
match  will  not  be  seriously  upset.  In  several 
high  power  transmitters  using  this  system,  and 
using  beam-tetrode  modulator  tubes,  it  is  pos- 
sible to  vary  the  plate  input  from  about  50 
watts  to  one  kilowatt  without  a change  other 
than  a slight  increase  in  audio  distortion  at 
the  adjustment  which  gives  the  lowest  power 
output  from  the  transmitter. 

With  triode  tubes  as  modulators  it  usually 
will  be  found  necessary  to  vary  the  grid  bias 
at  the  same  time  that  the  plate  voltage  is 
changed.  This  will  allow  the  tubes  to  be  op- 
erated at  approximately  the  same  relative  point 
on  their  operating  characteristic  when  the  plate 
voltage  is  varied.  When  the  modulator  tubes  are 
operated  with  zero  bias  at  full  plate  voltage,  it 
will  usually  be  possible  to  reduce  the  modu- 
lator voltage  along  with  the  voltage  on  the 
modulated  stage,  with  no  apparent  change  in 
the  voice  quality.  However,  it  will  be  necessary 
to  reduce  the  audio  gain  at  the  same  time  that 
the  plate  voltage  is  reduced. 

Transmitter 
Control  Methods 

Almost  everyone,  when  getting  a new  trans- 


392  Transmitter  Keying  and  Control 


THE  RADIO 


Figure  6 

TRANSMITTER  CONTROL  CIRCUIT 

Closing  Sj  lights  all  filaments  In  the  transmitter  and  starts  the  time-delay  relay  in  its  cycle. 
When  the  time-delay  relay  has  operated,  closing  the  transmit-receive  switch  at  the  operating  po- 
sition will  apply  plate  power  to  the  transmitter  and  disable  the  receiver.  A tune-up  switch  has 
been  provided  so  that  the  exciter  stages  may  be  tuned  without  plate  voltage  on  the  final 

amplifier. 


mitter  on  the  air,  has  had  the  experience  of 
having  to  throw  several  switches  and  pull  or 
insert  a few  plugs  when  changing  from  receive 
to  transmit.  This  is  one  extreme  in  the  direc- 
tion of  how  not  to  control  a transmitter.  At  the 
other  extreme  we  find  systems  where  it  is  only 
necessary  to  speak  into  the  microphone  or 
touch  the  key  to  change  both  transmitter  and 
receiver  over  to  the  transmit  condition.  Most 
amateur  stations  are  intermediate  between  the 
two  extremes  in  the  control  provisions  and  use 
some  relatively  simple  system  for  transmitter 
control. 

In  figure  5 is  shown  an  arrangement  which 
protects  mercury-vapor  rectifiers  against  pre- 
mature application  of  plate  voltage  without 
resorting  to  a time-delay  relay.  No  matter  which 
switch  is  thrown  first,  the  filaments  will  be 
turned  on  first  and  off  last.  However,  double- 
pole switches  are  required  in  place  of  the  usual 
single-pole  switches. 

When  assured  time  delay  of  the  proper  inter- 
val and  greater  operating  convenience  are  de- 
sired, a group  of  inexpensive  a-c  relays  may 


be  incorporated  into  the  circuit  to  give  a con- 
trol circuit  such  as  is  shown  in  figure  6.  This 
arrangement  uses  a 115-volt  thermal  (or  motor- 
operated)  time-delay  relay  and  a d-p-d-t  115- 
volt  control  relay.  Note  that  the  protective 
interlocks  are  connected  in  series  with  the 
coil  of  the  relay  which  applies  high  voltage  to 
the  transmitter.  A tune-up  switch  has  been  in- 
cluded so  that  the  transmittet  may  be  tuned  up 
as  far  as  the  grid  circuit  of  the  final  stage  is 
concerned  before  application  of  high  voltage 
to  the  final  amplifier.  Provisions  for  operat- 
ing an  antenna-changeover  relay  and  for  cut- 
ting the  plate  voltage  to  the  receiver  when  the 
transmitter  is  operating  have  been  included. 

A circuit  similar  to  that  of  figure  6 but  in- 
corporating push-button  control  of  the  trans- 
mitter is  shown  in  figure  7.  The  circuit  features 
a set  of  START-STOP  and  TRANSMIT-RE- 
CEIVE buttons  at  the  transmitter  and  a sepa- 
rate set  at  the  operating  position.  The  control 
push  buttons  operate  independently  so  that 
either  set  may  be  used  to  control  the  trans- 
mitter. It  is  only  necessary  to  push  the  START 


HANDBOOK 


Safety  Precautions  393 


ALL  FILAMENT  TRANSFORMERS  EXCITER  H.V.  HIGH  VOLTAGE 

TRANSFORMER  TRANSFORMER 


Figure  7 

PUSH-BUTTON  TRANSMITTER-CONTROL  CIRCUIT 

Pushing  the  START  button  either  at  the  transmitter  or  at  the  operating  position  will  light  all 
filaments  and  start  the  time-delay  relay  in  its  cycle.  When  the  cycle  has  been  completed,  a 
touch  of  the  TRANSMIT  button  will  put  the  transmitter  on  the  air  and  disable  the  receiver.  Push- 
ing the  RECEIVE  button  will  disable  the  transmitter  and  restore  the  receiver.  Pushing  the  STOP 
button  will  instantly  drop  the  entire  transmitter  from  the  a-c  line.  If  desired,  a switch  may  be 
placed  in  series  with  the  lead  from  the  RECEIVE  button  to  the  protective  interlocksi  opening 
the  switch  will  make  it  impossible  for  any  person  accidentally  to  put  the  tronsm/ffer  on  the  air. 
Various  other  safety  provisions,  such  as  the  protective-interlock  arrangement  described  in  the 

text  have  been  incorporated. 

With  the  circuit  arrangement  shown  for  the  overload-relay  contacts,  it  is  only  necessary  to  use 
a simple  normally-closed  d-c  relay  with  a varioble  shunt  across  the  coil  of  the  relay.  When  the 
current  through  the  coil  becomes  great  enough  to  open  the  normally-closed  contacts  the  hold- 
circuit  on  the  plate-voltage  relay  will  be  broken  and  the  plate  voltage  will  be  removed.  If  the 
overload  is  only  momentary,  suah  as  a modulation  peak  or  a tank  floshover,  merely  pushing  the 
TRANSMIT  button  will  again  put  the  transmitter  on  the  air.  This  simple  circuit  provision  elimi- 
nates the  requirement  for  expensive  overload  relays  of  the  mechanically-latching  type,  but  still 
gives  excellerrt  overload  protection. 


button  momentarily  to  light  the  transmitter  fila- 
ments and  start  the  time-delay  relay  in  its  cy- 
cle. When  the  standby  light  comes  on  it  is  only 
necessary  to  touch  the  TRANSMIT  button  to 
put  the  transmitter  on  the  air  and  disable  the 
receiver.  Touching  the  RECEIVE  button  will 
turn  off  the  transmitter  and  restore  the  receiver. 
After  a period  of  operation  it  is  only  necessary 
to  touch  the  STOP  button  at  either  the  trafls- 
mitter  or  the  operating  position  to  shut  down 
the  transmitter.  This  type  of  control  arrange- 
ment is  called  an  electrically-locking  push-to- 


transmit  control  system.  Such  systems  are  fre- 
quently used  in  industrial  electronic  control. 


20-3  Safety  Precautions 

The  best  way  for  an  operator  to  avoid  ser- 

ious accidents  from  the  high  voltage  supplies 

of  a transmitter  is  for  him  to  use  his  head,  act 
only  with  deliberation,  and  not  take  unneces- 


394 


Transmitter  Keying  and  Control 


THE  RADIO 


sary  chances.  However,  no  one  is  infallible, 
and  chances  of  an  accident  are  greatly  less- 
ened if  certain  factors  are  taken  into  consid- 
eration in  the  design  of  a transmitter,  in  order 
to  protect  the  operator  in  the  event  of  a lapse 
of  caution.  If  there  are  too  many  things  one 
must  "watch  out  for”  or  keep  in  mind  there  is 
a good  chance  that  sooner  or  later  there  will 
be  a mishap;  and  it  only  takes  one.  When  de- 
signing or  constructing  a transmitter,  the  fol- 
lowing safety  considerations  should  be  given 
attention. 

Grounds  For  the  utmost  in  protection,  every- 
thing of  metal  on  the  front  panel  of 
a transmitter  capable  of  being  touched  by  the 
operator  should  be  ar  ground  potenrial.  This 
includes  dial  set  screws,  meter  zero  adjuster 
screws,  merer  cases  if  of  metal,  meter  jacks, 
everything  of  metal  protruding  through  the  front 
panel  or  capable  of  being  touched  or  nearly 
touched  by  the  operator.  This  applies  whether 
or  not  the  panel  itself  is  of  metal.  Do  not  rely 
upon  the  insulation  of  meter  cases  or  tuning 
knobs  for  protection. 

The  B negative  or  chassis  of  all  plate  power 
supplies  should  be  connected  together,  and  to 
an  external  ground  such  as  a watetpipe. 

Exposed  Wires  It  is  not  necessary  to  resort 
and  Components  to  rack  and  panel  construc- 
tion in  order  to  provide  com- 
plete enclosure  of  all  components  and  wiring 
of  the  transmitter.  Even  with  metal-chassis 
construction  it  is  possible  to  arrange  things  so 
as  to  incorporate  a protective  shielding  hous- 
ing which  will  not  interfere  with  ventilation 
yet  will  prevent  contact  with  all  wires  and 
components  carrying  high  voltage  d.c.  or  a.c., 
in  addition  to  offering  shielding  action. 

If  everything  on  the  front  panel  is  at  ground 
potential  (with  respect  to  external  ground)  and 
all  units  are  effectively  housed  with  protective 
covers,  then  there  is  no  danger  except  when 
the  operator  must  reach  into  the  interior  part 
of  the  transmitter,  as  when  changing  coils, 
neutralizing,  adjusting  coupling,  or  shooting 
trouble.  The  latter  procedure  can  be  made  safe 
by  making  it  possible  for  the  operator  to  be 
absolutely  certain  that  all  voltages  have  been 
turned  off  and  that  they  cannot  be  turned  on 
either  by  short  circuit  or  accident.  This  can  be 
done  by  incorporation  of  the  following  system 
of  main  primary  switch  and  safety  signal  lights. 

Combined  Safety  The  common  method  of 
Signal  and  Switch  using  red  pilot  lights  to 
show  when  a circuit  is  on 
is  useless  except  from  an  ornamenral  srand- 
point.  When  rhe  red  pilor  is  not  lit  it  usually 
means  that  the  circuit  is  turned  off,  bur  ir  can 


6.3V.  TO  GREEN  PILOT  LIGHTS  ON 

FRONT  PANEL  AND  ON  EACH  CHASSIS 

000 

o o 

/ 

= FIL.  TRANS. 

N 

/ 

MAIN  IIS  V.  SUPPLY  <5~--Q  D.P.O.T.  SWITCH 

1 I 

115  V.A.C.  TO  ENTIRE  TRANSMITTER 

Figure  8 

COMBINED  MAIN  SWITCH  AND 
SAFETY  SIGNAL 

When  shutting  down  the  transmitter,  throw 
the  main  switch  to  neutrai.  If  work  is  to  be 
done  on  the  transmitter,  throw  the  switch  all 
the  way  to  "pilot,”  thus  turning  on  the  green 
pilot  lights  on  the  panel  and  on  each  chas- 
sis, and  insuring  that  no  voltage  can  exist 
on  the  primary  of  any  transformer,  even  by 
virtue  of  a short  or  accidental  ground. 


mean  thar  the  circuit  is  on  but  the  lamp  is 
burned  out  or  not  making  contact. 

To  enable  you  to  touch  the  tank  coils  in 
your  transmitter  with  absolute  assurance  rhar 
it  is  impossible  for  you  to  obtain  a shock  ex- 
cept from  possible  undischarged  filter  conden- 
sers (see  following  topic  for  elimination  of 
this  hazard),  ir  is  only  necessary  to  incorpo- 
rate a device  similar  to  that  of  figure  8.  It  is 
placed  neat  the  point  where  the  main  110- volt 
leads  enter  the  room  (preferably  near  the  door) 
and  in  such  a position  as  to  be  inaccessible 
to  small  children.  Notice  that  this  switch  breaks 
both  leads;  switches  that  open  just  one  lead 
do  not  afford  complete  protection,  as  it  is 
sometimes  possible  to  complete  a primary  cir- 
cuit through  a short  or  accidental  ground.  Break- 
ing just  one  side  of  the  line  may  be  all  tight 
for  turning  the  transmitter  on  and  off,  but  when 
you  are  going  to  place  an  atm  inside  the  trans- 
mitter, both  110-volt  leads  should  be  broken. 

When  you  ate  all  through  working  your  trans- 
mitter for  the  time  being,  simply  throw  the 
main  switch  to  neutral. 

When  you  find  it  necessary  to  work  on  the 
transmitter  or  change  coils,  throw  the  switch 
so  that  the  green  pilots  light  up.  These  can  be 
ordinary  6.3-volt  pilot  lamps  behind  green 
bezels  or  dipped  in  green  lacquer.  One  should 
be  placed  on  the  front  panel  of  the  transmitter; 
others  should  be  placed  so  as  to  be  easily 
visible  when  changing  coils  or  making  adjust- 
ments requiting  the  operator  to  reach  inside 
the  transmitter. 


HANDBOOK 


Transmitter  Keying  395 


For  100  per  cenr  protection,  just  obey  the 
following  rule:  never  work  on  the  transmitter 
or  reach  inside  any  protective  cover  except 
when  the  green  pilots  are  glowing.  To  avoid 
confusion,  no  other  green  pilots  should  be  used 
on  the  transmitter;  if  you  want  an  indicator 
jewel  to  show  when  the  filaments  ate  lighted, 
use  amber  instead  of  green. 

Safety  Bleeders  Filter  capacitors  of  good  qual- 
ity hold  their  charge  for  some 
time,  and  when  the  voltage  is  more  than  1000 
volts  it  is  just  about  as  dangerous  to  get  across 
an  undischarged  4-fifd.  filter  capacitor  as  it  is 
to  get  across  a high-voltage  supply  that  is 
turned  on.  Most  power  supplies  incorporate 
bleeders  to  improve  regulation,  but  as  these 
are  generally  wire-wound  resistors,  and  as 
wire- wound  resistors  occasionally  open  up 
without  apparent  cause,  it  is  desirable  to  in- 
corporate an  auxiliary  safety  bleeder  across 
each  heavy-duty  bleeder.  Carbon  resistors  will 
not  stand  much  dissipation  and  sometimes 
change  in  value  slightly  with  age.  However, 
the  chance  of  their  opening  up  when  run  well 
within  their  dissipation  rating  is  very  small. 

To  make  sure  that  all  capacitors  are  bled,  it 
is  best  to  short  each  one  with  an  insulated 
screwdriver.  However,  this  is  sometimes  awk- 
ward and  ahways  inconvenient.  One  can  be  vir- 
tually sure  by  connecting  auxiliary  carbon 
bleeders  across  all  wire-wound  bleeders  used 
on  supplies  of  1000  volts  or  more.  For  every 
500  volts,  connect  in  series  a 500,000-ohm 
1-watt  carbon  resistor.  The  drain  will  be  neg- 
ligible (1  ma.)  and  each  resistor  will  have  to 
dissipate  only  0.5  watt.  Under  these  condi- 
tions the  resistors  will  last  indefinitely  with 
little  chance  of  opening  up.  For  a 1500-volt 
supply,  connect  three  500,000-ohm  resistors  in 
series.  If  the  voltage  exceeds  an  integral  num- 
ber of  500  volt  divisions,  assume  it  is  the  next 
higher  integral  value;  for  instance,  assume 
1800  volts  as  2000  volts  and  use  four  resistors. 

Do  not  attempt  to  use  fewer  resistors  by 
using  a higher  value  for  the  resistors;  not  over 
500  volts  should  appear  across  any  single 
1-watt  resistor. 

In  the  event  that  the  regular  bleeder  opens 
up,  it  will  take  several  seconds  for  the  auxil- 
iary bleeder  to  drain  the  capacitors  down  to  a 
safe  voltage,  because  of  the  very  high  resist- 
ance. Therefore,  it  is  best  to  allow  10  or  15 
seconds  after  turning  off  the  plate  supply  be- 
fore attempting  to  work  on  the  ttansmitter. 

If  a 0-1  d-c  milliammeter  is  at  hand,  it  may 
be  connected  in  series  with  the  auxiliary 
bleeder  to  act  as  a high  voltage  voltmeter. 

“Hot”  Adjustments  Some  amateurs  contend 
that  it  is  almost  impossible 


to  make  certain  adjustments,  such  as  coupling 
and  neutralizing,  unless  the  transmitter  is  run- 
ning. The  best,  thing  to  do  is  to  make  all  neu- 
tralizing and  coupling  devices  adjustable  from 
the  front  panel  by  means  of  flexible  control 
shafts  which  are  broken  with  insulated  cou- 
plings to  permit  grounding  of  the  panel  beating. 

If  your  particular  transmitter  layout  is  such 
that  this  is  impracticable  and  you  refuse  to 
throw  the  main  switch  to  make  an  adjustment 
—throw  the  main  switch— take  a reading- throw 
the  main  switch— make  an  adjustment— and  so 
on,  then  protect  yourself  by  making  use  of  long 
adjusting  rods  made  from  i^-inch  dowel  sticks 
which  have  been  wiped  with  oil  when  perfectly 
free  from  moisture. 

If  you  are  addicted  to  the  use  of  pickup  loop 
and  flashlight  bulb  as  a resonance  and  neutral- 
izing indicator,  then  fasten  it  to  the  end  of  a 
long  dowel  stick  and  use  it  in  that  manner. 

Protective  Interlocks  With  the  increasing  ten- 
dency toward  construc- 
tion of  transmitters  in  enclosed  steel  cabinets 
a transmitter  becomes  a particularly  lethal  de- 
vice unless  adequate  safety  provisions  have 
been  incorporated.  Even  with  a combined  safety 
signal  and  switch  as  shown  in  figure  8 it  is 
still  conceivable  that  some  person  unfamiliar 
with  the  transmitter  could  come  in  contact  with 
high  voltage.  It  is  therefore  recommended  that 
the  transmitter,  wherever  possible,  be  built 
into  a complete  metal  housing  or  cabinet  and 
that  all  doors  or  access  covers  be  provided 
with  protective  interlocks  (all  interlocks  must 
be  connected  in  series)  to  remove  the  high 
voltage  whenever  these  doors  or  covers  are 
opened.  The  term  "high  voltage”  should  mean 
any  voltage  above  approximately  150  volts, 
although  it  is  still  possible  to  obtain  a serious 
burn  from  a 150-volt  circuit  under  certain  cir- 
cumstances. The  150-volt  limit  usually  will 
mean  that  grid-bias  packs  as  well  as  high- 
voltage  packs  should  have  their  primary  cir- 
cuits opened  when  any  interlock  is  opened. 


20-4  Transmitter  Keying 

The  carrier  from  a c-w  telegraph  transmitter 
must  be  broken  into  dots  and  dashes  for  the 
transmission  of  code  characters.  The  carrier 
signal  is  of  constant  amplitude  while  the  key 
is  closed,  and  is  entirely  removed  when  the 
key  is  open.  When  code  characters  are  being 
transmitted,  the  carrier  may  be  considered  as 
being  modulated  by  the  keying.  If  the  change 
from  the  no-output  condition  to  full-output,  or 
vice  versa,  occurs  too  rapidly,  the  rectangular 

pulses  which  form  the  keying  characters  con- 
tain high-frequency  components  which  take  up 


396  Transmitter  Keying  and  Control 


THE  RADIO 


a wide  frequency  band  as  sidebands  and  are 
heard  as  clicks. 

The  cure  for  transient  key  clicks  is  rela- 
tively simple,  although  one  would  not  believe 
it,  judging  from  the  hordes  of  dicky,  "snappy” 
signals  heard  on  the  ait. 

To  be  capable  of  transmitting  code  charac- 
ters and  at  the  same  time  not  splitting  the 
eardrums  of  neighboring  amateurs,  the  c-w 
transmitter  MUST  meet  two  important  speci- 
fications. 

1-  It  must  have  no  parasitic  oscillations 
either  in  the  stage  being  keyed  or  in  any 
succeeding  stage. 

2-  It  must  have  some  device  in  the  keying 
circuit  capable  of  shaping  the  leading 
and  trailing  edge  of  the  waveform. 

Both  these  specifications  must  he  met  be- 
fore the  transmitter  is  capable  of  c-w  opera- 
tion. Merely  turning  a transmitter  on  and  off 
by  the  haphazard  insertion  of  a telegraph  key 
in  some  power  lead  is  an  invitation  to  trouble. 

The  two  general  methods  of  keying  a trans- 
mitter are  those  which  control  the  excitation 
to  the  keyed  amplifier,  and  those  which  con- 
trol the  plate  or  screen  voltage  applied  to  the 
keyed  amplifier. 

Key-Click  Key-click  elimination  is  accom- 

Ellminatlon  plished  by  preventing  a too-rapid 

make-and-break  of  power  to  the 
antenna  circuit,  rounding  off  the  keying  char- 
acters so  as  to  limit  the  sidebands  to  a value 
which  does  not  cause  interference  to  adjacent 
channels.  Too  much  lag  will  prevent  fast  key- 
ing, but  fortunately  key  clicks  can  be  prac- 
tically eliminated  without  limiting  the  speed 
of  manual  (hand)  keying.  Some  circuits  which 
eliminate  key  clicks  introduce  too  much  time- 
lag  and  thereby  add  tails  to  the  dots.  These 
tails  may  cause  the  signals  to  be  difficult  to 
copy  at  high  speeds. 

Location  of  Considerable  thought  should  be 
Keyed  Stage  given  as  to  which  stage  in  a 
transmitter  is  the  proper  one  to 
key.  If  the  transmitter  is  keyed  in  a stage  close 
to  the  oscillator,  the  change  in  r-f  loading  of 
the  oscillator  will  cause  the  oscillator  to  shift 
frequency  with  keying.  This  will  cause  the 
signal  to  have  a distinct  chirp.  The  chirp  will 
be  multiplied  as  many  times  as  the  frequency 
of  the  oscillator  is  multiplied.  A chirpy  oscil- 
lator that  would  be  passable  on  80  meters 
would  be  unusable  on  28  Me.  c.w. 

Keying  the  oscillator  itself  is  an  excellent 
way  to  run  into  keying  difficulties.  If  no  key 
click  filter  is  used  in  the  keying  circuit,  the 
transmitter  will  have  bad  key  clicks.  If  a key 
click  filter  is  used,  the  slow  rise  and  decay 
of  oscillator  voltage  induced  by  the  filter  ac- 
tion will  cause  a keying  chirp.  This  action  is 


CENTER-TAP  KEYING  WITH  CLICK 
FILTER 

The  constants  shown  above  are  suggested  as 
starting  values;  considerable  variation  in 
these  values  can  be  expected  for  optimum 
keying  of  amplifiers  of  different  operating 
conditions,  it  is  suggested  that  a keying  re- 
lay be  substituted  for  the  key  in  the  circuit 
above  wherever  practicable. 


true  of  all  oscillators,  whether  electron  coupled 
or  crystal  controlled. 

The  more  amplifier  or  doubler  stages  that 
follow  the  keyed  stage,  the  more  difficult  it  is 
to  hold  control  of  the  shape  of  the  keyed  wave- 
form. A heavily  excited  doubler  stage  or  class 
C stage  acts  as  a peak  clipper,  tending  to 
square  up  a rounded  keying  impulse,  and  the 
cumulative  effect  of  several  such  stages  cas- 
caded is  sufficient  to  square  up  the  keyed 
waveform  to  the  point  where  bad  clicks  ate 
reimposed  on  a clean  signal. 

A good  rule  of  thumb  is  to  never  key  back 
farther  than  one  stage  removed  from  the  final 
amplifier  stage,  and  never  key  closer  than  one 
stage  removed  from  the  frequency  controlling 
oscillator  of  the  transmitter.  Thus  there  will 
always  be  one  isolating  stage  between  the 
keyed  stage  and  the  oscillator,  and  one  isolat- 
ing stage  between  the  keyed  stage  and  the 
antenna.  At  this  point  the  waveform  of  the 
keyed  signal  may  be  most  easily  controlled. 

Keyer  Circuit  In  the  first  place  it  may  be  es- 
Requirements  tablished  that  the  majority  of 
new  design  transmitters,  and 
many  of  those  of  older  design  as  well,  use  a 
medium  power  beam  tetrode  tube  either  as  the 
output  stage  or  as  the  exciter  for  the  output 
stage  of  a high  power  transmitter.  Thus  the 
transmitter  usually  will  end  up  with  a tube 
such  as  type  2E26,  807,  6146,  813,  4-65A, 
4E27/257B,  4- 125 A or  similar,  or  one  of  these 
tubes  will  be  used  as  the  stage  just  ahead  of 
the  output  stage. 

Second,  it  may  be  established  that  it  is  un- 
desirable to  key  further  down  in  the  transmitter 
chain  than  the  stage  just  ahead  of  the  final 


HANDBOOK 


Cathode  Keying  397 


f .ywi  • W 

VACUUM  TUBE  KEYERS  FOR  CENTER-TAP  KEYING  CIRCUITS 
r/ie  type  A keyer  is  suitable  for  keying  stages  running  up  to  1250  volts  on  the  plate.  Two  2A3 
or  6A3  tubes  can  safely  key  160  milliamperes  of  cathode  current.  The  simple  6Y6  keyer  in  fig- 
ure B IS  for  keying  stages  running  up  to  650  volts  on  the  plate.  A single  6Y6  con  key  80  milli- 
amperes. Two  in  parallel  may  be  used  for  plate  currents  under  160  ma.  If  softer  keying  is  de- 
sired, the  500-p.p.fd.  mica  condenser  should  be  Increased  to  .001  pfd. 


amplifier.  If  a low-level  stage,  which  is  fol- 
lowed by  a series  of  class  C amplifiers,  is 
keyed,  serious  transients  will  be  generated 
in  the  output  of  the  transmitter  even  though 
the  keyed  stage  is  being  turned  on  and  off  very 
smoothly.  This  condition  arises  as  a result  of 
pulse  sharpening,  which  has  been  discussed 
previously. 

Third,  the  output  from  the  stage  should  be 
completely  cut  off  when  the  key  is  up,  and  the 
time  constant  of  the  rise  and  decay  of  the  key- 
ing wave  should  be  easily  controllable. 

Fourth,  it  should  be  possible  to  make  the 
rise  period  and  the  decay  period  of  the  keying 
wave  approximately  equal.  This  type  of  keying 
envelope  is  the  only  one  tolerable  for  commer- 
cial work,  and  is  equally  desirable  for  obtain- 
ing clean  cut  and  easily  readable  signals  in 
amateur  work. 

Fifth,  it  is  desirable  that  the  keying  circuit 
be  usable  without  a keying  relay,  even  when 
a high-power  stage  is  being  keyed. 

Last,  for  the  sake  of  simplicity  and  safety, 
it  should  be  possible  to  ground  the  frame  of 
the  key,  and  yet  the  circuit  should  be  such 


that  placing  the  fingers  across  the  key  will 
notresultin  an  electrical  shock.  In  other  words, 
the  keying  circuit  should  be  inherently  safe. 

All  these  requirements  have  been  met  in  the 
keying  circuits  to  be  described. 


20-5  Cathode  Keying 

The  lead  from  the  cathode  or  center-tap  con- 
nection of  the  filament  of  an  r-f  amplifier  can 
be  opened  and  closed  for  a keying  circuit.  Such 
a keying  system  opens  the  plate  voltage  cir- 
cuit and  at  the  same  time  opens  the  grid  bias 
return  lead.  For  this  reason,  the  grid  circuit 
is  blocked  at  the  same  time  the  plate  circuit 
is  opened.  This  helps  to  reduce  the  backwave 
that  might  otherwise  leak  through  the  keyed 
stage. 

The  simplest  cathode  keying  circuit  is  il- 
lustrated  in  figure  9,  where  a key-click  filter 
is  employed,  and  a hand  key  is  used  to  break 
the  circuit.  This  »imple  keying  circuit  is  not 


3 98  Transmitter  Keying  and  Controi 


THE  RADIO 


-BLOCKING  +L.V.  +H.V. 

BIAS 


Figure  1 1 

SIMPLE  BLOCKED-GRID  KEYING 
SYSTEM 

The  blocking  bios  must  be  sufficient  to  cut- 
off plate  current  to  the  amplifier  stage  in  the 
presence  of  the  excitation  voltage-  Rj  is  nor- 
mal bias  resistor  for  the  tube.  R3  and  Cj 
should  be  adjusted  for  correct  keying  wave- 
form. 


HI-MU  TRIODE 


Figure  12 

SELF-BLOCKING  KEYING  SYSTEM  FOR 
HIGH-MU  TRIODE 

R]  and  Cj  adjusted  for  correct  keying  wave- 
form. Rj  is  bias  resistor  of  tube. 


recommended  for  general  use,  as  considerable 
voltage  will  be  developed  across  the  key  when 
it  is  open. 

An  electronic  switch  can  take  the  place  of 
the  hand  key.  This  will  remove  the  danger  of 
shock.  At  the  same  time,  the  opening  and  clos- 
ing characteristics  of  the  electronic  switch 
may  easily  be  altered  to  suit  the  particular 
need  at  hand.  Such  an  electronic  switch  is 
called  a vacuum  tube  keyer.  Low  internal  re- 
sistance triode  tubes  such  as  the  45,  6A3,  or 
6AS7  are  used  in  the  keyer.  These  tubes  act 
as  a very  high  resistance  when  sufficient 


807,  6 146,  ETC. 


Figure  13 

TWO-STAGE  BLOCKED-GRID  KEYER 

A separate  filament  transformer  must  be  used 
for  the  6J5,  as  its  filament  is  at  a potential 
of  —400  volts. 


blocking  bias  is  applied  to  them,  and  as  a 
very  low  resistance  when  the  bias  is  removed. 
The  desired  amount  of  lag  or  cushioning  effect 
can  be  obtained  by  employing  suitable  resist- 
ance and  capacitance  values  in  the  grid  of  the 
keyer  tube(s).  Because  very  little  spark  is 
produced  at  the  key,  due  to  the  small  amount 
of  power  in  the  key  circuit,  sparking  clicks 
are  easily  suppressed. 

One  type  45  tube  should  be  used  for  every 
50  ma.  of  plate  current.  Type  6B4G  or  2A3 
tubes  may  also  be  used;  allow  one  6B4G  tube 
for  every  80  ma.  of  plate  current. 

Because  of  the  series  resistance  of  the  keyer 
tubes,  the  plate  voltage  at  the  keyed  tube  will 
be  from  30  to  60  volts  less  than  the  power 
supply  voltage.  This  voltage  appears  as  cath- 
ode bias  on  the  keyed  tube,  assuming  the  bias 
return  is  made  to  ground,  and  should  be  taken 
into  consideration  when  providing  bias. 

Some  typical  cathode  circuit  vacuum  tube 
keying  units  are  shown  in  figure  10. 


20-6  Grid  Circuit  Keying 

Grid  circuit,  or  blocked  grid  keying  is  an- 
other effective  method  of  keying  a c-w  trans- 
mitter. A basic  blocked  grid  keying  circuit  is 
shown  in  figure  11.  The  time  constant  of  the 
keying  is  determined  by  the  RC  circuit,  which 
also  forms  part  of  the  bias  circuit  of  the  tube. 
Vifhen  the  key  is  closed,  operating  bias  is  de- 
veloped by  the  flow  of  grid  current  through  Rj 
When  the  key  is  open,  sufficient  fixed  bias  is 
applied  to  the  tube  to  block  it,  preventing  the 
stage  from  functioning.  If  an  un-neutralized 


HANDBOOK 


Screen  Grid  Keying  399 


807,  ETC. 


Figure  14 

SINGLE-STAGE  SCREEN  GRID  KEYER 
FOR  TETRODE  TUBES 


tetrode  is  keyed  by  this  method,  there  is  the 
possibility  of  a considerable  backwave  caused 
by  r-£  leakage  through  the  grid-plate  capacity 
of  the  tube. 

Certain  hi-/i  triode  tubes,  such  as  the  8 11- A 
and  the  805,  automatically  block  themselves 
when  the  grid  return  circuit  is  opened.  It  is 
merely  necessary  to  insert  a key  and  associ- 
ated key  click  filter  in  the  grid  return  lead  of 
these  tubes.  No  blocking  bias  supply  is  need- 
ed. This  circuit  is  shown  in  figure  12. 

A more  elaborate  blocked-grid  keying  sys- 
tem has  been  developed  by  WIDX,  and  was 
shown  in  the  February,  1954  issue  of  QST 
magazine.  This  highly  recommended  circuit 
is  shown  in  figure  13.  Two  stages  are  keyed. 


preventing  any  backwave  emission.  The  first 
keyed  stage  may  be  the  oscillator,  or  a low 
powered  buffer.  The  last  keyed  stage  may  be 
the  driver  stage  to  the  power  amplifier,  or  the 
amplifier  itself.  Since  the  circuit  is  so  pro- 
portioned that  the  lower  powered  stage  comes 
on  first  and  goes  off  last,  any  keying  chirp  in 
the  oscillator  is  not  emitted  on  the  air.  Keying 
lag  is  applied  ro  the  high  powered  keyed  stage 
only. 


20-7  Screen  Grid  Keying 

The  screen  circuit  of  a tetrode  tube  may  be 
keyed  for  c-w  operation.  Unfortunately,  when 
the  screen  grid  of  a terrode  tube  is  brought  to 
zero  potential,  the  tube  still  delivers  con- 
siderable output.  Thus  it  is  necessary  to  place 
a negative  blocking  voltage  on  the  screen  grid 
to  reduce  the  backwave  through  the  tube.  A 
suitable  keyer  circuit  that  will  achieve  this 
was  developed  by  W6DTY,  and  was  described 
in  the  February,  1953  issue  of  CQ  magazine. 
This  circuit  is  shown  in  figure  14.  A 6L6  is 
used  as  a combined  clamper  tube  and  keying 
tube.  When  the  key  is  closed,  the  6L6  tube 
has  blocking  bias  applied  to  its  control  grid. 
This  bias  is  obtained  from  the  rectified  grid 
bias  of  the  keyed  tube.  Screen  voltage  is  ap- 
plied to  the  keyed  stage  through  a screen  drop- 
ping resistor  and  a VR-105  regulator  tube. 
When  the  key  is  open,  the  6L6  is  no  longer 
cut-off,  and  conducts  heavily.  The  voltage 
drop  across  the  dropping  resistor  caused  by 
the  heavy  plate  current  of  the  6L6  lowers  the 
voltage  on  the  VR-105  tube  until  it  is  extin- 


Figure  15 

TOP  VIEW  OF  SCREEN 
GRID  KEYER  SHOWN  IN 
FIGURE  16 


400  Transmitter  Keying  and  Control 


THE  RADIO 


LOW  POWER  BUFFER 


guished,  removing  the  screen  voltage  from  the 
tetrode  r-f  tube.  At  the  same  time,  rectified 
grid  bias  is  applied  to  the  screen  of  the  tetrode 
through  the  1 megohm  resistor  between  screen 
and  key.  This  voltage  effectively  cuts  off  the 
screen  of  the  tetrode  until  the  key  is  closed 
again.  The  RC  circuit  in  the  grid  of  the  6L6 
tube  determines  the  keying  characteristic  of 
the  tetrode  tube. 

A more  elaborate  screen  grid  keyer  is  shown 
in  figures  15  and  16.  This  keyer  is  designed 
to  block-grid  key  the  oscillator  or  a low  pow- 
ered buffer  stage,  and  to  screen  key  a medium 
powered  tetrode  tube  such  as  an  807,  2E26  or 
6146.  The  unit  described  includes  a simple 
dual  voltage  power  supply  for  the  positive 
screen  voltage  of  the  tetrode,  and  a negative 
supply  for  the  keyer  stages.  A 6K6  is  used  as 
the  screen  keyer,  and  a 12AU7  is  used  as  a 
cathode  follower  and  grid  block  keyer.  As  in 


the  WIDX  keyer,  this  keyer  turns  on  the  ex- 
citer a moment  before  the  tetrode  stage  is 
turned  on.  The  tetrode  stage  goes  off  an  in- 
stant before  the  exciter  does.  Thus  any  key- 
ing chirp  of  the  oscillator  is  effectively  re- 
moved from  the  keyed  signal. 

By  listening  in  the  receiver  one  can  hear 
the  exciter  stop  operating  a fraction  of  a sec- 
ond after  the  tetrode  stage  goes  off.  In  fact, 
during  rapid  keying,  the  exciter  may  be  heard 
as  a steady  signal  in  the  receiver,  as  it  has 
appreciable  time  lag  in  the  keying  circuit.  The 
clipping  effect  of  following  stages  has  a defi- 
nite hardening  effect  on  this,  however. 

20-8  Differential  Keying  Circuits 

Excellent  waveshaping  may  be  obtained  by 
a differential  keying  system  whereby  the 
master  oscillator  of  the  transmitter  is  turned 


RELATIVE  OUTPUT 


HANDBOOK 


Differential  Keying  401 


Figure  18 


BLOCKING  DIODES  EMPLOYED 
TO  VARY  TIME  CONSTANT  OF 
“MAKE” AND “BREAK” CHARAC- 
TERISTICS OF  VACUUM  TUBE 
KEYER 


Figure  17 

TIME  SEQUENCE  OF  A 
DIFFERENTIAL  KEYER 


on  a moment  before  the  rest  of  the  stages  are 
energized,  and  remains  on  a moment  longer 
than  the  other  stages.  The  ^chirp”  or  frequen* 
cy  shift  associated  with  abrupt  switching  of 
the  oscillator  is  thus  removed  from  the  emitted 
signal.  In  addition,  the  differential  keyer  can 
apply  waveshaping  to  the  amplifier  section 
of  the  transmitter,  eliminating  the  "click” 
caused  by  rapid  keying  of  the  latter  stages. 

The  ideal  keying  system  would  perform  as 
illustrated  in  figure  17.  When  the  key  is 
closed,  the  oscillator  reaches  maximum  out- 
put almost  instantaneously.  The  following 
stages  reach  maximum  output  in  a fashion 
determined  by  the  waveshaping  circuits  of 
the  keyer.  When  the  key  is  released,  the  out- 
put of  the  amplifier  stages  starts  to  decay 
in  a predetermined  manner,  followed  shortly 
thereafter  by  cessation  of  the  oscillator.  The 
overall  result  of  these  actions  is  to  provide 
relatively  soft  "make”  and  "break”  to  the 
keyed  signal,  meanwhile  preventing  oscilla- 
tor frequency  shift  during  the  keying  se- 
quence. 

The  rates  of  charge  and  decay  in  a typical 
R-C  keying  circuit  may  be  varied  independent- 
ly of  each  other  by  the  blocking  diode  system 
of  figure  18.  Each  diode  permits  the  charging 
current  of  the  timing  capacitor  to  flow  through 
only  one  of  the  two  variable  potentiometers, 
thus  permitting  independent  adjustment  of 
the  "make”  and  "break”  characteristics  of 
the  keying  system. 

A practical  differential  keying  system  de- 


veloped by  WUCP  (Feb.,  1956  QST)  is  shown 
in  figure  19.  A 6AL5  switch  tube  turns  the 
oscillator  on  before  the  keying  action  starts, 
and  holds  it  on  until  after  the  keying  se- 
quence is  completed.  Time  constant  of  the 
keying  cycle  is  determined  by  values  of  C 
and  R.  When  the  key  is  open,  a cut-off  bias 
of  about  —110  volts  is  applied  to  the  screen 
grid  circuits  of  the  keyed  stages.  When  the 
key  is  closed,  the  screen  grid  voltage  rises 
to  the  normal  value  at  a rate  determined  by 
the  time  constant  R-C.  Upon  opening  the  key 
again,  the  screen  voltage  returns  to  cut-off 
value  at  the  predetermined  rate. 

The  potentiometer  R1  serves  as  an  output 
control,  varying  the  minimum  internal  re- 
sistance of  the  12BH7  keyer  tube,  and  is  a 
useful  device  to  limit  power  input  during  tune- 
up  periods.  Excitation  to  the  final  amplifier 
stage  may  be  controlled  by  the  screen  po- 
tentiometer R3  in  the  second  buffer  stage. 
An  external  bias  source  of  approximately  — 120 
volts  at  10  milliamperes  is  required  for  oper- 
ation of  the  keyer,  in  addition  to  the  300-volt 
screen  supply. 

Blocking  voltage  may  be  removed  from  the 
oscillator  for  "zeroing”  purposes  by  closing 
switch  SI,  rendering  the  diode  switch  in- 
operative. 

A second  popular  keying  system  is  shown 
in  figure  20,  and  is  widely  used  in  many 
Johnson  transmitters.  Grid  block  keying  is 
used  on  tubes  V2  and  V3.  A waveshaping 
filter  consisting  of  R2,  R3>  ^nd  Cl  is  used 
in  the  keying  control  circuit  of  V2  and  V3. 

To  avoid  chirp  when  the  oscillator  (VI)  is 

keyed,  the  keyer  tube  V4  allows  the  oscillator 
to  start  quickly — before  V2  and  V3  start 


402 


OSCILLATOR  BUFFER*!  BUFFER  #2 


- 1 20  V. 


Figure  19 

DIFFERENTIAL  KEYING  SYSTEM  WITH 
OSCILLATOR  SWITCHING  DIODE 


Vi  V2  V3 

OSCILLATOR  BUFFER  DRIVER 


Figure  20 

DIFFERENTIAL  KEYER  EMPLOYED  IN 
‘‘JOHNSON”  TRANSMITTERS 


conducting and  continue  operating  This  may  be  adjusted  to  cut  off  the  VFO 

until  after  V2  and  V3  have  stopped  con-  between  marks  of  keyed  characters,  thus 
ducting.  Potentiometer  Rl  adjusts  the  "hold”  allowing  rapid  break-in  operation, 
time  for  VFO  operation  after  the  key  is  opened. 


CHAPTER  TWENTY-ONE 


Radiation,  Propagation 
and  Transmission  Lines 


Radio  waves  are  electromagnetic  waves 
similar  in  nature  but  much  lower  in  frequency 
than  light  waves  or  heat  waves.  Such  waves 
represent  electric  energy  traveling  through 
space.  Radio  waves  travel  in  free  space  with 
the  velocity  of  light  and  can  be  reflected  and 
refracted  much  the  same  as  light  waves. 


21-1  Radiation  from  an 

Antenna 

Alternating  current  passing  through  a con- 
ductor creates  an  alternating  electromagnetic 
field  around  that  conductor.  Energy  is  alter- 
nately stored  in  the  field,  and  then  returned 
to  the  conductor.  As  the  frequency  is  raised, 
more  and  more  of  the  energy  does  not  return 
to  the  conductor,  but  instead  is  radiated  off 
into  space  in  the  form  of  electromagnetic 
waves,  called  radio  waves.  Radiation  from  a 
wire,  or  wires,  is  materially  increased  when- 
ever there  is  a sudden  change  in  the  electrical 
constants  of  the  line.  These  sudden  changes 
produce  reflection,  which  places  standing 
waves  on  the  line. 

When  a wire  in  space  is  fed  radio  frequency 
energy  having  a wavelength  of  approximately 

2. 1 times  the  length  of  the  wire  in  meters,  the 

wire  resonates  as  a half-wave  dipole  antenna 
at  that  wavelength  or  frequency.  The  greatest 


possible  change  in  the  electrical  constants 
of  a line  is  that  which  occurs  at  the  open  end 
of  a wire.  Therefore,  a dipole  has  a great  mis- 
match at  each  end,  producing  a high  degree  of 
reflection.  We  say  that  the  ends  of  a dipole 
are  terminated  in  an  infinite  impedance. 

A returning  wave  which  has  been  reflected 
meets  the  next  incident  wave,  and  the  voltage 
and  current  at  any  point  along  the  antenna  are 
the  vector  sum  of  the  two  waves.  At  the  ends 
of  the  dipole,  the  voltages  add,  while  the  cur- 
rents of  the  two  waves  cancel,  thus  producing 
high  voltage  and  low  current  at  the  ends  of  the 
dipole  or  half  wave  section  of  wire.  In  the 
same  manner,  it  is  found  that  the  currents  add 
while  the  voltages  cancel  at  the  center  of  the 
dipole.  Thus,  at  the  center  there  is  high  cur- 
rent but  low  voltage. 

Inspection  of  figure  1 will  show  that  the 
current  in  a dipole  decreases  sinusoidally 
towards  either  end,  while  the  voltage  similarly 
increases.  The  voltages  at  the  two  ends  of  the 
antenna  are  180°  out  of  phase,  which  means 
that  the  polarities  are  opposite,  one  being  plus 
while  the  other  is  minus  at  any  instant.  A 
curve  representing  either  the  voltage  or  cur- 
rent on  a dipole  represents  a standing  wave 
on  the  wire. 

Radiation  from  Radiation  can  and  does  take 

Sources  other  place  from  sources  other  than 

than  Antennos  antennas.  Undesired  radiation 
can  take  place  from  open-wire 


403 


4 04  Radiation,  Propagation  and  Lines 


THE  RADIO 


'vVOLTASe 


HALF-WAVE  ANTENNA - 


SHOWING  HOW  STANDING  WAVES 
EXIST  ON  A HORIZONTAL  ANTENNA. 


Figure  1 

STANDING  WAVES  ON  A RESONANT 
ANTENNA 


transmission  lines,  both  from  single-wire  lines 
and  from  lines  comprised  of  more  than  one 
wire.  In  addition,  radiation  can  be  made  to 
take  place  in  a very  efficient  manner  from  elec- 
tromagnetic horns,  from  plastic  lenses  or  from 
elecrromagnetic  lenses  made  up  of  spaced  con- 
ducting planes,  from  slots  cut  in  a piece  of 
metal,  from  dielectric  wires,  or  from  the  open 
end  of  a wave  guide. 


Directivity  of  The  radiation  from  any  phys- 

Rodiotion  ically  practicable  radiating 

system  is  directive  to  a certain 
degree.  The  degree  of  directivity  can  be  en- 
hanced or  altered  when  desirable  through  the 
combination  of  radiating  elements  in  a pre- 
scribed manner,  through  the  use  of  reflecting 
planes  or  curved  surfaces,  or  through  the  use 
of  such  systems  as  mentioned  in  the  preceding 
paragraph.  The  construction  of  directive  an- 
tenna arrays  is  covered  in  detail  in  the  chap- 
ters which  follow. 

Polarization  Like  light  waves,  radio  waves 
can  have  a definite  polarization. 
In  fact,  while  light  waves  ordinarily  have  to 
be  reflected  or  passed  through  a polarizing 
medium  before  they  have  a definite  polariza- 
tion, a radio  wave  leaving  a simple  radiator 
will  have  a definite  polarization,  the  polar- 
ization being  indicated  by  the  orientation  of 
the  electric-field  componenr  of  the  wave.  This, 
in  turn,  is  determined  by  the  orientation  of  the 
radiator  itself,  as  the  magnetic-field  component 
is  always  at  right  angles  to  a linear  radiator, 
and  the  electric-field  component  is  always  in 
the  same  plane  as  the  radiator.  Thus  we  see 
that  an  antenna  that  is  vertical  with  respect 
to  the  earth  will  transmit  a vertically  polar- 
ized wave,  as  the  electrostatic  lines  of  force 
will  be  vertical.  Likewise,  a simple  horizontal 
antenna  will  radiate  horizontally  polarized 
waves. 


Because  the  orientation  of  a simple  linear 

radiator  is  the  same  as  the  polarization  of  the 
waves  emitted  by  it,  tbe  radiator  itself  is  re- 
ferred to  as  being  either  vertically  or  horizon- 
tally polarized.  Thus,  we  say  thar  a horizontal 
antenna  is  horizontally  polarized. 

Figure  2A  illustrares  the  fact  that  the  polar- 
ization of  the  electric  field  of  the  radiation 
from  a vertical  dipole  is  vertical.  Figure  2B, 
on  the  other  hand,  shows  that  the  polarization 
of  electric-field  radiation  from  a vertical  slot 
radiator  is  horizontal.  This  fact  has  been  uti- 
lized in  certain  commercial  FM  antennas  where 
it  is  desired  to  have  horizontally  polarized 
radiation  but  where  it  is  more  convenient  to 
use  an  array  of  vertically  stacked  slot  arrays. 
If  the  metallic  sheet  is  bent  into  a cylinder 
with  the  slot  on  one  side,  substantially  omni- 
directional horizontal  coverage  is  obtained 
with  horizontally-polarized  radiation  when  the 
cylinder  with  the  slot  in  one  side  is  oriented 
vertically.  An  arrangement  of  this  type  is  shown 
in  figure  2C.  Several  such  cylinders  may  be 
stacked  vertically  to  reduce  high-angle  radia- 
tion and  to  concentrate  the  radiated  energy 
at  the  useful  low  radiation  angles. 

In  any  event  the  polarization  of  radiation 
from  a radiating  system  is  parallel  to  the  elec- 
tric field  as  it  is  set  up  inside  or  in  the  vici- 
nity of  the  radiating  system. 


21-2  General  Character- 

istics of  Antennas 

All  antennas  have  certain  general  character- 
istics to  be  enumerated.  It  is  the  result  of 
differences  in  these  general  characteristics 
which  makes  one  type  of  antenna  system  most 
suitable  for  one  type  of  application  and  an- 
other type  best  for  a different  application.  Six 
of  the  more  important  characteristics  are;  (1) 
polarization,  (2)  radiation  resistance,  (3)  hori- 
zontal directivity,  (4)  vertical  directivity, 
(5)  bandwidth,  and  (6)  effective  power  gain. 

The  polarization  of  an  antenna  or  radiating 
system  is  the  direction  of  the  electric  field 
and  has  been  defined  in  Section  21-1. 

The  radiation  resistance  of  an  antenna  sys- 
tem is  normally  referred  to  the  feed  point  in 
an  antenna  fed  at  a current  loop,  or  it  is  re- 
ferred to  a current  loop  in  an  antenna  system 
fed  at  another  point.  The  radiation  resistance 
is  that  value  of  resistance  which,  if  insetted 
in  series  with  the  antenna  at  a current  loop, 
would  dissipate  the  same  energy  as  is  actually 
radiated  by  the  antenna  if  the  antenna  current 
at  the  feed  point  were  to  remain  the  same. 

The  horizontal  and  vertical  directivity  can 
best  be  expressed  as  a directive  pattern  which 


HANDBOOK 


Antenna  Characteristics  405 


Figure  2 

ANTENNA  POLARIZATION 

The  polarization  (electric  field)  of 
the  radiation  from  a resonant  dipole 
such  as  shown  at  (A)  above  is  paral- 
lel to  the  length  of  the  radiator.  In 
the  case  of  a resonant  slot  cut  in  a 
sheet  of  metal  and  used  as  a radia- 
tor, the  polarization  (of  the  elec- 
tric field)  is  perpendicular  to  the 
length  of  the  slot.  In  both  cases, 
however,  the  polarization  of  the 
radiated  field  is  parallel  to  the  po- 
tential gradient  of  the  radiator;  in 
the  case  of  the  dipole  the  electric 
lines  of  force  are  from  end  to  end, 
while  in  the  ease  of  the  slot  the 
field  is  across  the  sides  of  the 
slot.  The  metallic  sheet  containing 
the  slot  may  be  formed  into  a cyl- 
inder to  make  up  the  radiator  shown 
at  (C).  With  this  type  of  radiator 
the  radiated  field  will  be  horizon- 
tally polarized  even  though  the 
radiator  is  mounted  vertically. 


is  a graph  showing  the  relative  radiated  field 
intensity  against  azimuth  angle  for  horizontal 
directivity  and  field  intensity  against  elevation 
angle  for  vertical  directivity. 

The  bandwidth  of  an  antenna  is  a measure 
of  its  ability  to  operate  within  specified  limits 
over  a range  of  frequencies.  Bandwidth  can 
be  expressed  either  ‘^operating  frequency  plus- 
or-minus  a specified  per  cent  of  operating  fre- 
quency** or  ‘‘operating  frequency  plus-or-minus 
a specified  number  of  megacycles'*  for  a cer- 
tain standing-wave-ratio  limit  on  the  trans- 
mission line  feeding  the  antenna  system. 

The  effective  power  gain  or  directive  gain 
of  an  antenna  is  the  ratio  between  the  power 
required  in  the  specified  antenna  and  the  power 
required  in  a reference  antenna  (usually  a half- 
wave dipole)  to  attain  the  same  field  strength 
in  the  favored  direction  of  the  antenna  under 
measurement.  Directive  gain  may  be  expressed 
either  as  an  actual  power  ratio,  or  as  is  more 
common,  the  power  ratio  may  be  expressed 
in  decibels. 

Physical  Length  If  the  cross  section  of  the 
of  a Half-Wave  conductor  which  makes  up 

Antenna  the  antenna  is  kept  very 

small  with  respect  to  the 
antenna  length,  an  electrical  half  wave  is  a 
fixed  percentage  shorter  than  a physical  half- 
wavelength. This  percentage  is  approximately 
5 per  cent.  Therefore,  most  linear  half-wave  an- 
tennas are  close  to  95  per  cent  of  a half  wave- 
length long  physically.  Thus,  a half-wave  an- 
tenna resonant  at  exactly  80  meters  would  be 
one-half  of  0.95  times  80  meters  in  length.  An- 
other way  of  saying  the  same  thing  is  that  a 


wire  resonates  at  a wavelength  of  about  2.1 
times  its  length  in  meters.  If  the  diameter  of 
the  conductor  begins  to  be  an  appreciable  frac- 
tion of  a wavelength,  as  when  tubing  is  used 
as  a v-h-f  radiator,  the  factor  becomes  slightly 
less  than  0.95*  For  the  use  of  wire  and  not 
tubing  on  frequencies  below  30  Me.,  however, 
the  figure  of  0.95  may  be  taken  as  accurate. 
This  assumes  a radiator  removed  from  sur- 
rounding objects,  and  with  no  bends. 

Simple  conversion  into  feet  can  be  obtained 
by  using  the  factor  1.56.  To  find  the  physical 
length  of  a half-wave  80-meter  antenna,  we 
multiply  80  times  1.56,  and  get  124.8  feet  for 
the  length  of  the  radiator. 

It  is  more  common  to  use  frequency  than 
wavelength  when  indicating  a specific  spot  in 
the  radio  spectrum.  For  this  reason,  the  rela- 
tionship between  wavelength  and  frequency 
must  be  kept  in  mind.  As  the  velocity  of  radio 
waves  through  space  is  constant  at  the  speed 
of  light,  it  will  be  seen  that  the  more  waves 
that  pass  a point  per  second  (higher  frequency), 
the  closer  together  the  peaks  of  those  waves 
must  be  (shorter  wavelength).  Therefore,  the 
higher  the  frequency,  the  lower  will  be  the 
wavelength. 

A radio  wave  in  space  can  be  compared  to 
a wave  in  water.  The  wave,  in  either  case,  has 
peaks  and  troughs.  One  peak  and  one  trough 
constitute  a full  wave,  or  one  wavelength. 

Frequency  describes  the  number  of  wave 
cycles  or  peaks  passing  a point  per  second. 
Wavelength  describes  the  distance  the  wave 
travels  through  space  during  one  cycle  or 
oscillation  of  the  antenna  current;  it  is  the 


4 06  Radiation,  Propagation  and  Lines 


THE  RADIO 


distance  in  meters  between  adjacent  peaks 

or  adjacent  troughs  of  a wave  train. 

As  a radio  wave  travels  300,000,000  meters 
a second  (speed  of  light),  a frequency  of  1 
cycle  per  second  corresponds  to  a wavelength 
of  300,000,000  meters.  So,  if  the  frequency  is 
multiplied  by  a million,  the  wavelength  must 
be  divided  by  a million,  in  order  to  maintain 
their  correct  ratio. 

A frequency  of  1,000,000  cycles  per  second 
(1,000  kc.)  equals  a wavelength  of  300  meters. 
Multiplying  frequency  by  10  and  dividing  wave- 
length by  10,  we  find:  a frequency  of  10,000  kc. 
equals  a wavelength  of  30  meters.  Multiplying 
and  dividing  by  10  again,  we  get:  a frequency 
of  100,000  kc.  equals  3 meters  wavelength. 
Therefore,  to  change  wavelength  to  frequency 
(in  kilocycles),  simply  divide  300,000  by  the 
wavelength  in  meters  (A). 

300,000 

Fkc  = ' - ■ 

A 

300,000 
X = — - ■'  ■'!  »■ 

Fkc 

Now  that  we  have  a simple  conversion  for- 
mula for  converting  wavelength  to  frequency 
and  vice  versa,  we  can  combine  it  with  our 
wavelength  versus  antenna  length  formula,  and 
we  have  the  following: 

Length  of  a half-wave  radiator  made  from 
wire  (no.  14  to  no.  10): 

3.5-Mc.  to  30-Mc.  bands 
468 

Length  in  feet  = 

Freq.  in  Me. 


50-Mc.  band 


Length  in  feet  = 
Length  in  inches  = 


460 


Freq.  in  Me. 
5600 

Freq.  in  Me. 


144-Mc.  band 

5500 

Length  in  inches  = 

Freq.  in  Me. 

Lenglh-to-Diameter  When  a half-wave  radiator 
Ratio  is  constructed  from  tubing 

or  rod  whose  diameter  is 
an  appreciable  fraction  of  the  length  of  the 
radiator,  the  resonant  length  of  a half-wave 
antenna  will  be  shortened.  The  amount  of 


Figure  3 

CHART  SHOWING  SHORTENING  OF  A 
RESONANT  ELEMENT  IN  TERMS  OF 
RATIO  OF  LENGTH  TO  DIAMETER 

The  use  of  this  chart  is  based  an  the  basic 
formula  where  radiator  length  in  feet  is 
equal  to  466/ frequency  in  Me.  This  formula 
applies  to  frequencies  below  perhaps  30  Me. 
when  the  radiator  is  made  from  wire.  On 
higher  frequencies,  or  on  14  and  28  Me.  when 
the  radiatar  is  made  of  large^diameter  tubing, 
the  radiator  is  shortened  from  the  value  ob- 
tained  with  the  above  formula  by  an  amount 
determined  by  the  ratio  of  length  to  diameter 
of  the  radiator.  The  amount  of  this  shorten^ 
ing  is  obtainable  from  the  chart  shown  above. 


shortening  can  be  determined  with  the  aid  of 
the  chart  of  figure  3.  In  this  chart  the  amount 
of  additional  shortening  over  the  values  given 
in  the  previous  paragraph  is  plotted  against 
the  ratio  of  the  length  to  the  diameter  of  the 
half-wave  radiator. 

The  length  of  a wave  in  free  space  is  some- 
what longer  than  the  length  of  an  antenna  for 
the  same  frequency.  The  actual  free-space 
half-wavelength  is  given  by  the  following 
expressions: 


492 

Half-wavelength  *=  — in  feet 

Freq.  in  Me. 

H If  , u 5905 

Half-wavelength  = in  inches 

Freq.  in  Me. 

Harmonic  A wire  in  space  can  resonate  at 
Rasononce  mote  than  one  frequency.  The  low- 
est frequency  at  which  it  resonates 
is  called  its  fundamental  frequency,  and  at 
that  frequency  it  is  approximately  a half  wave- 
length long.  A wire  can  have  two,  three,  four, 
five,  or  more  standing  waves  on  it,  and  thus 
it  resonates  at  approximately  the  integral  har- 
monics of  its  fundamental  frequency.  However, 
the  higher  harmonics  are  not  exactly  integral 
multiples  of  the  lowest  resonant  frequency  as 
a result  of  end  effects. 


HANDBOOK 


Radiation  Resistance  407 


A harmonic  operated  antenna  is  somewhat 
longer  than  the  corresponding  integral  number 
of  dipoles,  and  for  this  reason,  the  dipole 
length  formula  cannot  be  used  simply  by  mul- 
tiplying by  the  corresponding  harmonic.  The 
intermediate  half  wave  sections  do  not  have 
end  effects.  Also,  the  current  distribution  is 
disturbed  by  the  fact  that  power  can  reach 
some  of  the  half  wave  sections  only  by  flowing 
through  other  sections,  the  latter  then  acting 
not  only  as  radiators,  but  also  as  transmission 
lines.  For  the  latter  reason,  the  resonant  length 
will  be  dependent  to  an  extent  upon  the  method 
of  feed,  as  there  will  be  less  attenuation  of 
the  current  along  the  antenna  if  it  is  fed  at  or 
near  the  center  than  if  fed  towards  or  at  one 
end.  Thus,  the  antenna  would  have  to  be  some- 
what longer  if  fed  near  one  end  than  if  fed  neat 
the  center.  The  difference  would  be  small, 
however,  unless  the  antenna  were  many  wave- 
lengths long. 

The  length  of  a center  fed  harmonically  oper- 
ated doublet  may  be  found  from  the  formula: 

^ _ (K-.05)  X 492 
Freq.  in  Me. 

where  K = number  of  waves  on 
antenna 

L = length  in  feet 

Under  conditions  of  severe  current  attenua- 
tion, it  is  possible  for  some  of  the  nodes,  or 
loops,  actually  to  be  slightly  greater  than  a 
physical  half  wavelength  apart.  Practice  has 
shown  that  the  most  practietd  method  of  reson- 
ating a harmonically  operated  antenna  accu- 
rately is  by  cut  and  try,  or  by  using  a feed 
system  in  which  both  the  feed  line  and  antenna 
ate  resonated  at  the  station  end  as  an  integral 
system. 

A dipole  or  half-wave  antenna  is  said  to 
operate  on  its  fundamental  or  first  harmonic. 
A full  wave  antenna,  1 wavelength  long,  oper- 
ates on  its  second  harmonic.  An  antenna  with 
five  half-wavelengths  on  it  would  be  operating 
on  its  fifth  harmonic.  Observe  that  the  fifth 
harmonic  antenna  is  wavelengths  long,  not 
5 wavelengths. 

Antenna  Most  types  of  antennas  operate 
Resonance  most  efficiently  when  tuned  or 
resonated  to  the  frequency  of 
operation.  This  consideration  of  course  does 
not  apply  to  the  rhombic  antenna  and  to  the 
parasitic  elements  of  arrays  employing  para- 
sitically  excited  elements.  However,  in  practi- 
cally every  other  case  it  will  be  found  that  in- 
creased efficiency  results  when  the  entire  an- 
tenna system  is  resonant,  whether  it  be  a sim- 
ple dipole  or  an  elaborate  array.  The  radiation 
efficiency  of  a resonant  wire  is  many  times 
that  of  a wire  which  is  not  resonant. 




X 

il 

Da  



Figure  4 

EFFECT  OF  SERIES  INDUCTANCE  AND 
CAPACITANCE  ON  THE  LENGTH  OF  A 
HALF-WAVE  RADIATOR 

The  top  antenna  has  been  electrically  length- 
ened by  placing  a coil  In  series  with  the  cen- 
ter. In  other  words,  on  antenna  with  a lumped 
inductance  in  its  center  can  be  made  shorter 
for  a given  frequency  than  a plain  wire  radia- 
tor. The  bottom  antenna  has  been  capacitive- 
ly  shortened  electrically.  In  other  words,  an 
antenna  with  a capacitor  in  series  with  it 
must  be  made  longer  for  a given  frequency 
since  its  effective  electrical  length  as  com- 
pared to  plain  wire  is  shorter. 


If  an  antenna  is  slightly  too  long,  it  can  be 
resonated  by  series  insertion  of  a variable 
capacitor  at  a high  current  point.  If  it  is  slight- 
ly too  short,  it  can  be  resonated  by  means  of 
a variable  inductance.  These  two  methods, 
illustrated  schematically  in  figure  4,  are  gen- 
erally employed  when  part  of  the  antenna  is 
brought  into  the  operating  room. 

With  an  antenna  array,  or  an  antenna  fed  by 
means  of  a transmission  line,  it  is  mote  com- 
mon to  cut  the  elements  to  exact  resonant 
length  by  "cut  and  try”  procedure.  Exact  an- 
tenna resonance  is  more  important  when  the 
antenna  system  has  low  radiation  resistance; 
an  antenna  with  low  radiation  resistance  has 
higher  Q (tunes  sharper)  than  an  antenna  with 
high  radiation  resistance.  The  higher  Q does 
not  indicate  greater  efficiency;  it  simply  indi- 
cates a sharper  resonance  curve. 

21-3  Radiation  Resistance 

and  Feed-Point  Impedance 

In  many  ways,  a half-wave  antenna  is  like  a 
tuned  tank  circuit.  The  main  difference  lies  in 
the  fact  that  the  elements  of  inductance,  capac- 
itance, and  resistance  ate  lumped  in  the  tank 
circuit,  and  are  distributed  throughout  the 
length  of  an  antenna.  The  center  of  a half-wave 
radiator  is  effectively  at  ground  potential  as 
far  as  r-f  voltage  is  concerned,  although  the 
current  is  highest  at  that  point. 


4 08  Radiation,  Propagation  and  Lines 


THE  RADIO 


10000 

9000 

aooo 

5 

Q 7000 

z ' 

y 6000' 

{/)  5000| 

to 

^ 1 

Loi*meter= 

1 

■■1 

1 

Hi 

1 

— 1 

— 1 

IIHI 



Bi 

\~ 

z 

o 

0.  3000 

Q 

UJ 

u 

U.  2000 

1000 

0 

■11 

imm 

r 

imm 

■■H 

■■■ 

L 

wmii 

Hi 

L 

|J^DIAMErE(i=-^  / 

l\Hi 

R 



rV, 

Q.I5X  0.5X  I.OA  I.5X  2,0X  2.5X 

OVERALL  LENGTH  OF  RADIATOR 


Figure  5 

FEED  POINT  RESISTANCE  OF  A CENTER 
DRIVEN  RADIATOR  AS  A FUNCTION  OF 
PHYSICAL  LENGTH  IN  TERMS  OF  FREE 
SPACE  WAVELENGTH 


When  the  antenna  is  resonant,  and  it  always 
should  be  for  best  results,  the  impedance  at 
the  center  is  substantially  resistive,  and  is 
termed  the  radiation  resistance.  Radiation  re- 
sistance is  a fictitious  term;  it  is  that  value 
of  resistance  (referred  to  the  current  loop) 
which  would  dissipate  the  same  amount  of 
power  as  being  radiated  by  the  antenna,  when 
fed  with  the  current  flowing  at  the  current  loop. 

The  radiation  resistance  depends  on  the 
antenna  length  and  its  proximity  to  nearby 
objects  which  either  absorb  or  re-radiate  pow- 
er, such  as  the  ground,  other  wires,  etc. 


Figure  6 

REACTIVE  COMPONENT  OF  THE  FEED 
POINT  IMPEDANCE  OF  A CENTER 
DRIVEN  RADIATOR  AS  A FUNCTION  OF 
PHYSICAL  LENGTH  IN  TERMS  OF  FREE 
SPACE  WAVELENGTH 


tenna  is  simply  one-half  of  a dipole.  For  that 
reason,  the  radiation  resistance  is  roughly 
half  the  73-ohm  impedance  of  the  dipole  or 
36.5  ohms.  The  radiation  resistance  of  a Mar- 
coni antenna  such  as  a mobile  whip  will  be 
lowered  by  the  proximity  of  the  automobile 
body. 


The  Marconi  Before  going  too  far  with  the 
Antenna  discussion  of  radiation  resist- 

ance, an  explanation  of  the  Mar- 
coni (grounded  quarter  wave)  anrenna  is  in 
order.  The  Marconi  antenna  is  a special  type 
of  Hertz  antenna  in  which  the  earth  acts  as  the 
"other  half”  of  the  dipole.  In  other  words,  the 
current  flows  into  the  earth  instead  of  into  a 
similar  quarter-wave  section.  Thus,  the  current 
loop  of  a Marconi  antenna  is  at  the  base  rather 
than  in  the  center.  In  either  case  it  is  a quartet 
wavelength  from  the  end. 

A half-wave  dipole  far  from  ground  and  other 
reflecting  objects  has  a radiation  resistance 
at  the  center  of  about  73  ohms.  A Marconi  an- 


Antenna  Because  the  power  throughout  the 
Impedance  antenna  is  the  same,  the  imped- 
ance of  a resonant  antenna  at  any 
point  along  its  length  merely  expresses  the 
ratio  between  voltage  and  current  at  that  point. 
Thus,  the  lowest  impedance  occurs  where  the 
current  is  highest,  namely,  at  the  center  of  a 
dipole,  or  a quarter  wave  from  the  end  of  a 
Marconi.  The  impedance  rises  uniformly  toward 
each  end,  where  it  is  about  2000  ohms  for  a 
dipole  remote  from  ground,  and  about  twice  as 
high  for  a vertical  Marconi. 

If  a vertical  half-wave  antenna  is  set  up  so 
that  its  lower  end  is  at  the  ground  level,  the 
effect  of  the  ground  reflection  is  to  increase 


HANDBOOK 


Antenna  impedance  409 


HEIGHT  IN  WAVELENGTHS  OF  CENTER  OF  VERTICAL 
HALF-WAVE  ANTENNA  ABOVE  PERFECT  GROUND 
.25  .3  .4  .5  .6  .7  .75 


HEIGHT  IN  WAVELENGTHS  OF  HORIZONTAL  HALF- 
WAVE ANTENNA  ABOVE  PERFECT  GROUND 


Figure  7 

EFFECT  OF  HEIGHT  ON  THE  RADIATION 
RESISTANCE  OF  A DIPOLE  SUSPENDED 
ABOVE  PERFECT  GROUND 


the  radiation  resistance  to  approximately  100 
ohms.  When  a horizontal  half-wave  antenna  is 
used,  the  radiation  resistance  (and,  of  course, 
the  amount  of  energy  radiated  for  a given  an- 
tenna current)  depends  on  the  height  of  the 
antenna  above  ground,  since  the  height  deter- 
mines the  phase  and  amplitude  of  the  wave 
reflected  from  the  ground  back  to  the  antenna. 
Thus  the  resultant  current  in  the  antenna  for 
a given  power  is  a function  of  antenna  height. 

Center-fed  When  a linear  radiator  is  series  fed 
Feed  Point  at  the  center,  the  resistive  and 
Impedance  reactive  components  of  the  driving 
point  impedance  are  dependent  up- 
on both  the  length  and  diameter  of  the  radiator 
in  wavelengths.  The  manner  in  which  the  resis- 
tive component  varies  with  the  physical  dimen- 
sions of  the  radiator  is  illustrated  in  figure  5. 
The  manner  in  which  the  reactive  component 
varies  is  illustrated  in  figure  6. 

Several  interesting  things  will  be  noted  with 
respect  to  these  curves.  The  reactive  com- 
ponent disappears  when  the  overall  physical 
length  is  slightly  less  than  any  number  of  half 
waves  long,  the  differential  increasing  with 
conductor  diameter.  For  overall  lengths  in  the 
vicinity  of  an  odd  number  of  half  wavelengths, 
the  center  feed  point  looks  to  the  generator  or 
transmission  line  like  a series-resonant  lumped 
circuit,  while  for  overall  lengths  in  the  vici- 
nity of  an  even  number  of  half  wavelengths,  it 
looks  like  a parallel-resonant  or  anti-resonant 
lumped  circuit.  Both  the  feed  point  resistance 


and  the  feed  point  reactance  change  more  slow- 
ly with  overall  radiator  length  (or  with  fre- 
quency with  a fixed  length)  as  the  conductor 
diameter  is  increased,  indicating  that  the  ef- 
fective "Q”  is  lowered  as  the  diameter  is  in- 
creased. However,  in  view  of  the  fact  that  the 
damping  resistance  is  neatly  all  "radiation 
resistance”  rather  than  loss  resistance,  the 
lower  Q does  not  represent  lower  efficiency. 
Therefore,  the  lower  Q is  desirable,  because 
it  permits  use  of  the  radiator  over  a wider  fre- 
quency range  without  resorting  to  means  for 
eliminating  the  reactive  component.  Thus,  the 
use  of  a large  diameter  conductor  makes  the 
overall  system  less  frequency  sensitive.  If  the 
diameter  is  made  sufficiently  large  in  terms  of 
wavelengths,  the  Q will  be  low  enough  to  qual- 
ify the  radiator  as  a "broad-band”  antenna. 

Tbe  curves  of  figure  7 indicate  the  theoreti- 
cal center-point  radiation  resistance  of  a half- 
wave antenna  for  various  heights  above  perfect 
ground.  These  values  are  of  importance  in 
matching  untuned  radio-frequency  feeders  to 
the  antenna,  in  order  to  obtain  a good  imped- 
ance match  and  an  absence  of  standing  waves 
on  the  feeders. 


Ground  Losses  Above  average  ground,  the 
actual  radiation  resistance 
of  a dipole  will  vary  from  the  exact  value  of 
figure  7 since  the  latter  assumes  a hypothet- 
ical, perfect  ground  having  no  loss  and  perfect 
reflection.  Fortunately,  the  curves  for  the  radi- 
ation resistance  over  most  types  of  earth  will 
correspond  rather  closely  with  those  of  the 
chart,  except  that  the  radiation  resistance  for 
a horizontal  dipole  does  not  fall  off  as  rapidly 
as  is  indicated  for  heights  below  an  eighth 
wavelength.  However,  with  the  antenna  so 
close  to  the  ground  and  the  soil  in  a strong 
field,  much  of  the  radiation  resistance  is  ac- 
tually represented  by  ground  loss;  this  means 
that  a good  portion  of  the  antenna  power  is 
being  dissipated  in  the  earth,  which,  unlike 
the  hypothetical  perfect  ground,  has  resistance. 
In  this  case,  an  appreciable  portion  of  the 
radiation  resistance  actually  is  loss  resist- 
ance. The  type  of  soil  also  has  an  effect  upon 
the  radiation  pattern,  especially  in  the  vertical 
plane,  as  will  be  seen  later. 

The  radiation  resistance  of  an  antenna  gen- 
erally increases  with  length,  although  this  in- 
crease varies  up  and  down  about  a constantly 
increasing  average.  The  peaks  and  dips  are 
caused  by  the  reactance  of  the  antenna,  when 
its  length  does  not  allow  it  to  resonate  at  the 
operating  frequency. 

Antenna  Antennas  have  a certain  loss  re- 
Efficiency  sistance  as  well  as  a radiation  re- 
sistance. The  loss  resistance  de- 
fines the  power  lost  in  the  antenna  due  to  ohm- 


41  0 Radiation,  Propagation  and  Lines 


THE  RADIO 


ic  resistance  of  the  wire,  ground  resistance 
(in  the  case  of  a Marconi),  corona  discharge, 
and  insulator  losses. 

The  approximate  effective  radiation  effi- 
ciency (expressed  as  a decimal)  is  equal  to; 
Nr  = Ra/(Ra+  where  is  equal  to  the 

radiation  resistance  and  Rl  is  equal  to  the 
effective  loss  resistance  of  the  antenna.  The 
loss  resistance  will  be  of  the  order  of  0.25 
ohm  for  large-diameter  tubing  conductors  such 
as  are  most  commonly  used  in  multi-element 
parasitic  arrays,  and  will  be  of  the  order  of 
0.5  to  2.0  ohms  for  arrays  of  normal  construc- 
tion using  copper  wire. 

When  the  radiation  resistance  of  an  antenna 
or  array  is  very  low,  the  current  at  a voltage 
node  will  be  quite  high  for  a given  power.  Like- 
wise, the  voltage  at  a current  node  will  be  very 
high.  Even  with  a heavy  conductor  and  excel- 
lent insulation,  the  losses  due  to  the  high  volt- 
age and  current  will  be  appreciable  if  the  radi- 
ation resistance  is  sufficiently  low. 

Usually,  it  is  not  considered  desirable  to 
use  an  antenna  or  array  with  a radiation  resist- 
ance of  less  than  approximately  5 ohms  unless 
there  is  sufficient  directivity,  compactness, 
or  other  advantage  to  offset  the  losses  result- 
ing from  the  low  radiation  resistance. 

Ground  The  radiation  resistance  of  a Mar- 
Resistance  coni  antenna,  especially,  should 
be  kept  as  high  as  possible.  This 
will  reduce  the  antenna  current  for  a given 
power,  thus  minimizing  loss  resulting  from  the 
series  resistance  offered  by  the  earth  connec- 
tion. The  radiation  resistance  can  be  kept  high 
by  making  the  Marconi  radiator  somewhat  longer 
than  a quarter  wave,  and  shortening  it  by  series 
capacitance  to  an  electrical  quarter  wave.  This 
reduces  the  current  flowing  in  the  earth  con- 
nection. It  also  should  be  removed  from  ground 
as  much  as  possible  (vertical  being  ideal). 
Methods  of  minimizing  the  resistance  of  the 
earth  connection  will  be  found  in  the  discus- 
sion of  the  Marconi  antenna. 

21-4  Antenna  Directivity 

All  practical  antennas  radiate  better  in  some 
directions  than  others.  This  characteristic  is 
called  directivity.  The  more  directive  an  an- 
tenna is,  the  more  it  concentrates  the  radiation 
in  a certain  direction,  or  directions.  The  more 
the  radiation  is  concentrated  in  a certain  direc- 
tion, the  greater  will  be  the  field  strength  pro- 
duced in  that  direction  for  a given  amount  of 
total  radiated  power.  Thus  the  use  of  a direc- 
tional antenna  or  array  produces  the  same  re- 
sult in  the  favored  direction  as  an  increase  in 
the  power  of  the  transmitter. 

The  increase  in  radiated  power  in  a certain 


direction  with  respect  to  an  antenna  in  free 
space  as  a result  of  inherent  directivity  is 
called  the  free  space  directivity  power  gain 
or  just  space  directivity  gain  of  the  antenna 
(referred  to  a hypothetical  isotropic  radiator 
which  is  assumed  to  radiate  equally  well  in 
all  directions).  Because  the  fictitious  isotropic 
radiator  is  a purely  academic  antenna,  not  phy- 
sically realizable,  it  is  common  practice  to  use 
as  a reference  antenna  the  simplest  unground- 
ed resonant  radiator,  the  half-wave  Hertz,  or 
resonant  doublet.  As  a half-wave  doublet  has 
a space  directivity  gain  of  2.15  db  over  an  iso- 
tropic radiator,  the  use  of  a resonant  dipole 
as  the  comparison  antenna  reduces  the  gain 
figure  of  an  array  by  2.15  db.  However,  it  should 
be  understood  that  power  gain  can  be  expressed 
with  regard  to  any  antenna,  just  so  long  as  it 
is  specified. 

As  a matter  of  interest,  the  directivity  of 
an  infinitesimal  dipole  provides  a free  space 
directivity  power  gain  of  1.5  (or  1.76  db)  over 
an  isotropic  radiator.  This  means  that  in  the 
direction  of  maximum  radiation  the  infinitesi- 
mal dipole  will  produce  the  same  field  of 
strength  as  an  isotropic  radiator  which  is  radi- 
ating 1.5  times  as  much  total  power. 

A half-wave  resonant  doublet,  because  of 
its  different  current  distribution  and  signifi- 
cant length,  exhibits  slightly  more  free  space 
power  gain  as  a result  of  directivity  than  does 
the  infinitesimal  dipole,  for  reasons  which  will 
be  explained  in  a later  section.  The  space 
directivity  power  gain  of  a half-wave  resonant 
doublet  is  1.63  (or  2.15  db)  referred  to  an  iso- 
tropic radiator. 

Horizontal  When  choosing  and  orienting  an 
Directivity  antenna  system,  the  radiation  pat- 
terns of  the  various  common  types 
of  antennas  should  be  given  careful  considera- 
tion. The  directional  charactetistics  are  of 
still  greater  importance  when  a directive  an- 
tenna array  is  used. 

Horizontal  directivity  is  always  desirable 
on  any  frequency  for  point-to-point  work.  How- 
ever, it  is  not  always  attainable  with  reason- 
able antenna  dimensions  on  the  lower  fre- 
quencies. Further,  when  it  is  attainable,  as 
on  the  frequencies  above  perhaps  7 Me.,  with 
reasonable  antenna  dimensions,  operating  con- 
venience is  greatly  furthered  if  the  maximum 
lobe  of  the  horizontal  directivity  is  control- 
lable. It  is  for  this  reason  that  rotatable  an- 
tenna arrays  have  come  into  such  common 
usage. 

Considerable  horizontal  directivity  can  be 
used  to  advantage  when:  (1)  only  point-to- 
point  work  is  necessary,  (2)  several  arrays  are 
available  so  that  directivity  may  be  changed 
by  selecting  or  reversing  antennas,  (3)  a single 
rotatable  array  is  in  use.  Signals  follow  the 


HANDBOOK 


Antenna  Directivity  411 


Figure  8 

VERTICAL-PLANE  DIRECTIONAL  CHAR- 
ACTERISTICS OF  HORIZONTAL  AND  VER- 
TICAL DOUBLETS  ELEVATED  0.6  WAVE- 
LENGTH AND  ABOVE  TWO  TYPES  OF 
GROUND 

H]  represents  a horizontal  doublet  over  typi- 
cal farmland,  H2  over  salt  water,  Vi  is  a 
vertical  pattern  of  radiation  from  a vertical 
doublet  over  typical  farmland^  Vj  over  salt 
water.  A salt  water  ground  is  the  closest 
approach  to  an  extensive  ideally  perfect 
ground  that  will  be  met  in  actual  practice. 


great-circle  path,  or  within  2 or  3 degrees  of 
that  path  under  all  normal  propagation  condi- 
tions. However,  under  turbulent  ionosphere 
conditions,  or  when  unusual  propagation  con- 
ditions exist,  the  deviation  from  the  great-circle 
path  for  greatest  signal  intensity  may  be  as 
great  as  90°.  Making  the  array  rotatable  over- 
comes these  difficulties,  but  arrays  having  ex- 
tremely high  horizontal  directivity  become  too 
cumbersome  to  be  rotated,  except  perhaps 
when  designed  for  operation  on  frequencies 
above  50  Me. 

Vertical  Vertical  directivity  is  of  the  great- 
Directivity  est  importance  in  obtaining  satis- 
factory communication  above  14 
Me.  whether  or  not  horizontal  directivity  is 
used.  This  is  true  simply  because  only  the 
energy  radiated  between  certain  definite  eleva- 
tion angles  is  useful  for  communication.  Ener- 


gy radiated  at  other  elevation  angles  is  lost 
and  performs  no  useful  function. 

Optimum  Angle  The  optimum  angle  of  radiation 
of  Radiation  for  propagation  of  signals  be- 
tween two  points  is  dependent 
upon  a number  of  variables.  Among  these  sig- 
nificant variables  are:  (1)  height  of  the  iono- 
sphere layer  which  is  providing  the  reflection, 
(2)  distance  between  the  two  stations,  (3)  num- 
ber of  hops  for  propagation  between  the  two 
stations.  For  communication  on  the  14-Mc. 
band  it  is  often  possible  for  different  modes 
of  propagation  to  provide  signals  between  two 
points.  This  means,  of  course,  that  more  than 
one  angle  of  radiation  can  be  used.  If  no  eleva- 
tion directivity  is  being  used  under  this  con- 
dition of  propagation,  selective  fading  will 
take  place  because  of  interference  between  the 
waves  arriving  over  the  different  paths. 

On  the  28-Mc.  band  it  is  by  fat  the  most  com- 
mon condition  that  only  one  mode  of  propagation 
will  be  possible  between  two  points  at  any 
one  time.  This  explains,  of  course,  the  reason 
why  rapid  fading  in  general  and  selective  fad- 
ing in  particular  are  almost  absent  from  sig- 
nals heard  on  the  28-Mc.  band  (except  for  fad- 
ing caused  by  local  effects). 

Measurements  have  shown  that  the  angles 
useful  for  communication  on  the  14-Mc.  band 
are  from  3°  to  about  30°;  angles  above  about 
15°  being  useful  only  for  local  work.  On  the 
28-Mc.  band  measurements  have  shown  that 
the  useful  angles  range  from  about  3°  to  18°; 
angles  above  about  12°  being  useful  only  for 
local  (less  than  3000  miles)  work.  These  fig- 
ures assume  normal  propagation  by  virtue  of 
the  layer. 

Angle  of  Radiation  It  now  becomes  of  inter- 
of  Typical  Antennas  est  to  determine  the  a- 
and  Arrays  mount  of  radiation  avail- 

able at  these  useful  low- 
er angles  of  radiation  from  commonly  used  an- 
tennas and  antenna  arrays.  Figure  8 shows 
relative  output  voltage  plotted  against  eleva- 
tion angle  (wave  angle)  in  degrees  above  the 
horizontal,  for  horizontal  and  vertical  doublets 
elevated  0.6  wavelength  above  two  types  of 
ground.  It  is  obvious  by  inspection  of  the 
curves  that  a horizontal  dipole  mounted  at  this 
height  above  ground  (20  feet  on  the  28-Mc. 
band)  is  radiating  only  a small  amount  of  ener- 
gy at  angles  useful  for  communication  on  the 
28-Mc.  band.  Most  of  the  energy  is  being  radi- 
ated uselessly  upward.  The  vertical  antenna 
above  a good  reflecting  surface  appears  much 
better  in  this  respect— and  this  fact  has  been 
proven  many  times  by  actual  installations. 

It  might  immediately  be  thought  that  the  a- 
mount  of  radiation  from  a horizontal  or  vertical 


412  Radiation,  Propagation  and  Lines 


THE  RADIO 


POWER  OUTPUT 


Figure  9 

VERTICAL  RADIATION 
PATTERNS 

Showing  the  vertical  radiation 
patterns  for  half~wave  antennas 
(or  co/inear  half-wave  or  ex- 
tended half-wave  antennas)  at 
different  heights  above  average 
ground  and  perfect  ground.  Note 
that  such  antennas  one-quarter 
wave  above  ground  concentrate 
most  radiation  at  the  very  high 
angles  which  are  useful  for  com- 
munication only  on  the  lower  fre- 
quency bands.  Antennas  one-half 
wave  above  ground  are  not 
shown,  but  the  elevation  pattern 
shows  one  lobe  on  each  side  at 
an  angle  of  30°  above  hori- 
zontal. 


dipole  could  be  increased  by  raising  the  an- 
tenna higher  above  the  ground.  This  is  true  to 
an  extent  in  the  case  of  the  horizontal  dipole; 
the  low-angle  radiation  does  increase  slowly 
after  a height  of  0.6  wavelength  is  reached 
but  at  the  expense  of  greatly  increased  high- 
angle  radiation  and  the  fotmation  of  a number 
of  nulls  in  the  elevation  pattern.  No  signal 
can  be  transmitted  or  received  at  the  elevation 
angles  where  these  nulls  have  been  formed. 
Tests  have  shown  that  a center  height  of  0.6 
wavelength  for  a vertical  dipole  (0.35  wave- 
length to  the  bottom  end)  is  about  optimum  for 
this  type  of  array. 

Figure  9 shows  the  effect  of  placing  a hori- 
zontal dipole  at  various  heights  above  ground. 
It  is  easily  seen  by  reference  to  figure  9 (and 
figure  10  which  shows  the  radiation  from  a di- 
pole at  wave  height)  that  a large  percentage 
of  the  total  radiation  from  the  dipole  is  being 
radiated  at  relatively  high  angles  which  are 
useless  for  communication  on  the  14-Mc.  and 
28-Mc.  bands.  Thus  we  see  that  in  order  to  ob- 
tain a worthwhile  increase  in  the  ratio  of  low- 
angle  radiation  to  high-angle  radiation  it  is 
necessary  to  place  the  antenna  high  above 
ground,  and  in  addition  it  is  necessary  to  use 


additional  means  for  suppressing  high-angle 
radiation. 

Suppression  of  High-angle  radiation  can  be 
High-ongle  suppressed,  and  this  radiation 
Radiation  can  be  added  to  that  going  out 

at  low  angles,  only  through  the 
use  of  some  sort  of  directive  antenna  system. 
There  are  three  general  types  of  antenna  ar- 
rays composed  of  dipole  elements  commonly 
used  which  concentrate  radiation  at  the  lower 
more  effective  angles  for  high-frequency  com- 
munication. These  types  are;  (1)  The  close- 
spaced out-of-phase  system  as  exemplified  by 
the  "flat-top”  beam  or  W8JK  array.  Such  con- 
figurations are  classified  as  end  fire  arrays. 

(2)  The  wide-spaced  in-phase  arrays,  as  exem- 
plified by  the  "Lazy  H”  antenna.  These  con- 
figurations are  classified  as  broadside  arrays. 

(3)  The  close-spaced  parasitic  systems,  as 
exemplified  by  the  three  element  rotary  beam. 

A comparison  between  the  radiation  from  a 
dipole,  a "flat-top  beam”  and  a pair  of  dipoles 
stacked  one  above  the  other  (half  of  a "lazy 
H”),  in  each  case  with  the  top  of  the  antenna 
at  a height  of  wavelength  is  shown  in  figure 
11.  The  improvement  in  the  amplitude  of  low- 


Figure  10 

VERTICAL  RADIATION 
PATTERNS 

Showing  vertical-plane  radiation 
patterns  of  a horizontal  single- 
section  flat-top  beam  with  one- 
eighth  wave  spacing  (solid 
curves)  and  a horizontal  half- 
wave anfenno  (dashed  curves) 
when  both  are  0.5  wavelength 
(A)  and  0.75  wavelength  (B)  a- 
bove  ground. 


GAIN  IN  FIELD  STRENGTH 


HANDBOOK 


Antenna  Bandwidth  413 


Figure  1 1 

COMPARATIVE  VERTICAL 
RADIATION  PATTERNS 

Showing  the  vertical  radiation 
patterns  of  a horizontal  single- 
section  flat-top  beam  (A),  an 
array  of  two  stacked  horizontal 
in-phase  half-wave  e/emen/s— 
half  of  a "Lazy  and 

a horizontal  dipole  (C),  In  each 
ease  the  top  of  the  antenna  sys- 
tem is  0.75  wavelength  above 
ground/,  as  shown  to  the  left  of 
the  curves. 


•5  1.0  1.5  £.0  2.5  3.0 


angle  radiation  at  the  expense  of  the  useless 
high-angle  radiation  with  these  simple  arrays 
as  contrasted  to  the  dipole  is  quite  marked. 

Figure  12  compares  the  patterns  of  a 3 ele- 
ment beam  and  a dipole  radiator  at  a height  of 
0.75  wavelength.  It  will  be  noticed  that  al- 
though there  is  more  energy  in  the  lobe  of  the 
beam  as  compared  to  the  dipole,  the  axis  of 
the  beam  is  at  the  same  angle  above  the  hori- 
zontal. Thus,  although  more  radiated  energy 
is  provided  by  the  beam  at  low  angles,  the 
average  angle  of  radiation  of  the  beam  is  no 
lower  than  the  average  angle  of  radiation  of 
the  dipole. 


21-5  Bandwidth 

The  bandwidth  of  an  antenna  or  an  antenna 
array  is  a function  primarily  of  the  radiation 
resistance  and  of  the  shape  of  the  conductors 
which  make  up  the  antenna  system.  For  arrays 
of  essentially  similar  construction  the  band- 
width (or  the  deviation  in  frequency  which  the 
system  can  handle  without  mismatch)  is  in- 
creased with  increasing  radiation  resistance, 
and  the  bandwidth  is  increased  with  the  use 
of  conductors  of  larger  diameter  (smaller  ratio 
of  length  to  diameter).  This  is  to  say  that  if 
an  array  of  any  type  is  constructed  of  large 
diameter  tubing  or  spaced  wires,  its  bandwidth 
will  be  greater  than  that  of  a similar  array 
constructed  of  single  wires. 

The  radiation  resistance  of  antenna  arrays 
of  the  types  mentioned  in  the  previous  para- 
graphs may  be  increased  through  the  use  of 
wider  spacing  between  elements.  With  increased 
radiation  resistance  in  such  arrays  the  radi~ 
ation  efficiency  increases  since  the  ohmic 
losses  within  the  conductors  become  a smaller 
percentage  of  the  radiation  resistance,  and  the 
bandwidth  is  increased  proportionately. 


21-6  Propagation  of 

Radio  Waves 

The  preceding  sections  have  discussed  the 
manner  in  which  an  electromagnetic-wave  or 
radio-wave  field  may  be  set  up  by  a radiating 
system.  However,  for  this  field  to  be  useful 
for  communication  it  must  be  propagated  to 
some  distant  point  where  it  may  be  received, 
or  where  it  may  be  reflected  so  that  it  may  be 
received  at  some  other  point.  Radio  waves 
may  be  propagated  to  a remote  point  by  either 
or  both  of  two  general  methods.  Propagation 


Figure  12 

VERTICAL  RADIATION  PATTERNS 

Showing  vertical  radiation  patterns  of  a hori- 
zontal dipole  (A)  and  a horizontal  3-element 
parasitic  array  (B)  of  o height  above  ground 
of  0,75  wavelength.  Note  that  the  axis  of  the 
main  radiation  lobes  are  at  the  same  angle 
above  the  horizontal.  Note  also  the  suppres- 
sion of  high  angle  radiation  by  the  parasitic 
array. 


414  Radiation,  Propagation  and  Lines 


the  radio 


Figure  13 

GROUND-WAVE  SIGNAL  PROPAGATION 

The  illustration  above  shows  the  three  com- 
ponents of  the  ground  wave:  tA),  the  surface 
wave;  (Bj,  the  direct  wave;  and  (C)^  the 
ground-reflected  wave.  The  direct  wave  and 
the  ground-reflected  wave  combine  at  the 
receiving  antenna  to  make  up  the  space 
wave. 


may  take  place  as  a result  of  the  ground  wave, 
or  as  a result  of  the  sky  wave  or  ionospheric 
wave. 

The  Ground  Wove  The  term  ground  wave  actu- 
ally includes  several  dif- 
ferent types  of  waves  which  usually  ate  called: 
(1)  the  surface  wave,  (2)  the  direct  wave,  and 
(3)  the  ground-reflected  wave.  The  latter  two 
waves  combine  at  the  receiving  antenna  to 
form  the  resultant  wave  or  the  space  wave. 
The  distinguishing  characteristic  of  the  com- 
ponents of  the  ground  wave  is  that  all  travel 
along  or  over  the  surface  of  the  earth,  so  that 
they  are  affected  by  the  conductivity  and  ter- 
rain of  the  earth’s  surface. 

The  Ionospheric  Wove  Intense  bombardment  of 
or  Sky  Wove  the  upper  regions  of 

the  atmosphere  by  radi- 
ations from  the  sun  results  in  the  formation 
of  ionized  layers.  These  ionized  layers,  which 
form  the  ionosphere,  have  the  capability  of 
reflecting  or  refracting  radio  waves  which  im- 
pinge upon  them.  A radio  wave  which  has  been 
propagated  as  a result  of  one  or  more  reflec- 
tions from  the  ionosphere  is  known  as  an 
ionospheric  wave  or  a sky  wave.  Such  waves 
make  possible  long  distance  radio  communica- 
tion. Propagation  of  radio  signals  by  iono- 
spheric waves  is  discussed  in  detail  in  Sec- 
tion 21-8. 

21-7  Ground-Wave 

Communication 

As  stated  in  the  preceding  paragraph,  the 
term  ground  wave  applies  both  to  the  surface 
wave  and  to  the  space  wave  (the  resultant 
wave  from  the  combination  of  the  direct  wave 
and  the  ground-reflected  wave)  or  to  a com- 


bination of  the  two.  The  three  waves  which 
may  combine  to  make  up  the  ground  wave  are 
illustrated  in  figure  13. 

The  Surface  Wave  The  surface  wave  is  that 
wave  which  we  normally 
receive  from  a standard  broadcast  station.  It 
travels  directly  along  the  ground  and  termin- 
ates on  the  earth’s  surface.  Since  the  earth  is 
a relatively  poor  conductor,  the  surface  wave 
is  attenuated  quite  rapidly.  The  surface  wave 
is  attenuated  less  rapidly  as  it  passes  over 
sea  water,  and  the  attenuation  decreases  for 
a specific  distance  as  the  frequency  is  de- 
creased. The  rate  of  attenuation  with  distance 
becomes  so  large  as  the  frequency  is  increased 
above  about  3 Me.  that  the  surface  wave  be- 
comes of  little  value  for  communication. 

The  Space  Wave  The  resultant  wave  or  space 
wave  is  illustrated  in  figure 
13  by  the  combination  of  (B)  and  (C).  It  is  this 
wave  path,  which  consists  of  the  combination 
of  the  direct  wave  and  the  ground-reflected 
wave  at  the  receiving  antenna,  which  is  the 
normal  path  of  signal  propagation  for  line-of- 
sight  or  near  line-of-sight  communication  or 
FM  and  TV  reception  on  frequencies  above 
about  40  Me. 

Below  line-of-sight  over  plane  earth  or 
water,  when  the  signal  source  is  effectively 
at  the  horizon,  the  ground-reflected  wave  does 
not  exist,  so  that  the  direct  wave  is  the  only 
component  which  goes  to  make  up  the  space 
wave.  But  when  both  the  signal  source  and  the 
receiving  antenna  ate  elevated  with  respect  to 
the  intervening  terrain,  the  ground-reflected 
wave  is  present  and  adds  vectorially  to  the 
direct  wave  at  the  receiving  antenna.  The  vec- 
torial addition  of  the  two  waves,  which  travel 
over  different  path  lengths  (since  one  of  the 
waves  has  been  reflected  from  the  ground)  re- 
sults in  an  interference  pattern.  The  interfer- 
ence between  the  two  waves  brings  about  a 
cyclic  variation  in  signal  strength  as  the  re- 
ceiving antenna  is  raised  above  the  ground. 
This  effect  is  illustrated  in  figure  14.  From 
this  figure  it  can  be  seen  that  best  space- 
wave  reception  of  a v-h-f  signal  often  will  be 
obtained  with  the  receiving  antenna  quite  close 
to  the  ground.  This  subject,  along  with  other 
aspects  of  v-h-f  signal  propagation  and  recep- 
tion, are  discussed  in  considerable  detail  in 
a book  on  fringe-area  TV  reception.* 

The  distance  from  an  elevated  point  to  the 
geometrical  horizon  is  given  by  the  approxi- 
mate equation:  d = 1.22\jW where 'the  distance 


* "Bettor  TV  Reception,’*  by  W.  W.  Smith  and  R.  L.  Daw- 
ley.  published  by  Editors  ond  Engineers,  Ltd.,  Summer- 
land,  Calif, 


HANDBOOK 


Ground  Wave  Communication  415 


WAVE  INTERFERENCE  WITH  HEIGHT 

When  the  source  of  a horixontally-polarized 
space-wave  signal  is  above  the  horizon,  the 
received  signal  at  a distant  location  will  go 
through  a cyclic  variation  as  the  antenna 
height  is  progressively  raised.  This  is  due 
to  the  difference  in  total  path  length  between 
the  direct  wave  and  the  ground-reflected 
wave,  and  to  the  fact  that  this  path  length 
difference  changes  with  antenna  height. 
When  the  path  length  difference  is  such  that 
the  two  waves  arrive  at  the  receiving  anten- 
na with  a phase  difference  of  360°  or  some 
multiple  of  360°,  the  two  waves  will  appear 
to  be  in  phase  as  far  as  the  antenna  is  con- 
cerned and  maximum  signal  will  be  obtained. 
On  the  other  hand,  when  the  antenna  height 
is  such  that  the  path  length  difference  for 
the  two  waves  causes  the  waves  to  arrive 
with  a phase  difference  of  an  odd  multiple 
of  180°  the  two  waves  will  substantially  can- 
cel, and  a null  will  be  obtained  at  that  an- 
tenna height.  The  difference  between  Dj 
and  D2  plus  D3  is  the  path-length  difference. 
Note  also  that  there  is  an  additional  180° 
phase  shift  in  the  graund-reflected  wave  at 
the  paint  where  it  is  reflected  from  the 
ground.  It  Is  this  latter  phase  shift  which 
causes  the  space-wave  field  intensity  af  a 
harizontally  polarized  wave  to  be  zero  with 
the  receiving  antenna  at  ground  level. 


d is  in  miles  and  the  antenna  height  H is  in 
feet.  This  equation  must  be  applied  separately 
to  the  transmitting  and  receiving  antennas  and 
the  results  added.  However,  refraction  and 
diffraction  of  the  signal  around  the  spherical 
earth  cause  a smaller  reduction  in  field  strength 
than  would  occur  in  the  absence  of  such  bend- 
ing, so  that  the  average  radio  horizon  is  some- 
what beyond  the  geometrical  horizon.  The 
equation  d = 1.4  \/h  is  sometimes  used  for 
determining  the  radio  horizon. 

Tropospheric  Propagation  by  signal  bending 

Propagation  in  the  lower  atmosphere,  called 

tropospheric  propagation,  can 
result  in  the  reception  of  signals  over  a much 
greater  distance  than  would  be  the  case  if  the 
lower  atmosphere  were  homogeneous.  In  a 
homogeneous  or  well-mixed  lower  atmosphere, 
called  a normal  or  standard  atmosphere,  there 
is  a gradual  and  uniform  decrease  in  index  of 
refraction  with  height.  This  effect  is  due  to 


the  combined  effects  of  a decrease  in  temper- 
ature, pressure,  and  water-vapor  content  with 
height. 

This  gradual  decrease  in  refractive  index 
with  height  causes  waves  radiated  at  very  low 
angles  with  respect  to  the  horizontal  to  be 
bent  downward  slightly  in  a curved  path.  The 
result  of  this  effect  is  that  such  waves  will  be 
propagated  beyond  the  true  or  geometrical 
horizon.  In  a so-called  standard  atmosphere 
the  effect  of  the  curved  path  is  the  same  as 
though  the  radius  of  the  earth  were  increased 
by  approximately  one  third.  This  condition  ex- 
tends the  horizon  by  approximately  30  per  cent 
for  normal  propagation,  and  the  extended- hori- 
zon is  known  as  the  radio  path  horizon,  men- 
tioned before. 

Conditions  Leading  to  When  the  temperature. 

Tropospheric  pressure,  or  water-vapor 

Stratification  content  of  the  atmos- 

phere does  not  change 
smoothly  with  rising  altitude,  the  discontinuity 
or  stratification  will  result  in  the  reflection 
or  refraction  of  incident  v-h-f  signals.  Ordi- 
narily this  condition  is  mote  prevalent  at  night 
and  in  the  summer.  In  certain  areas,  such  as 
along  the  west  coast  of  North  America,  it  is 
frequent  enough  to  be  considered  normal.  Sig- 
nal strength  decreases  slowly  with  distance 
and,  if  the  favorable  condition  in  the  lower 
atmosphere  covets  sufficient  area,  the  range 
is  limited  only  by  the  transmitter  power,  an- 
tenna gain,  receiver  sensitivity,  and  signal-to- 
noise  ratio.  There  is  no  skip  distance.  Usually, 
transmission  due  to  this  condition  is  accom- 
panied by  slow  fading,  although  fading  can  be 
violent  at  a point  where  direct  waves  of  about 
the  same  strength  are  also  received. 

Bending  in  the  troposphere,  which  refers  to 
the  region  from  the  earth’s  surface  up  to  about 
10  kilometers,  is  more  likely  to  occur  on  days 
when  there  are  stratus  clouds  than  on  clear, 
cool  days  with  a deep  blue  sky.  The  tempera- 
ture or  humidity  discontinuities  may  be  broken 
up  by  vertical  convection  currents  over  land 
in  the  daytime  but  are  more  likely  to  continue 
during  the  day  over  water.  This  condition  is 
in  some  degree  predictable  from  weather  infor- 
mation several  days  in  advance.  It  does  not 
depend  on  the  sunspot  cycle.  Like  direct  com- 
munication, best  results  require  similar  an- 
tenna polarization  or  orientation  at  both  the 
transmitting  and  receiving  ends,  whereas  in 
transmission  via  reflection  in  the  ionosphere 
(that  part  of  the  atmosphere  between  about  50 
and  500  kilometers  high)  it  makes  little  dif- 
ference whether  antennas  are  similarly  polar- 
ized. 

Duct  Formation  When  bending  conditions  are 
particularly  favorable  they 


416  Radiation,  Propagation  and  Lines 


THE  RADIO 


Figure  15 

ILLUSTRATING  DUCT  TYPES 

Showing  two  types  of  variation  in  refractive 
index  with  height  which  will  give  rise  to  the 
formation  of  a duct.  An  elevated  duct  is 
shown  at  (A),  and  a ground-based  duct  is 
shown  at  (B),  Such  ducts  can  propagate 
ground-wave  signals  far  beyond  their  normal 
range. 


may  give  rise  to  the  formation  of  a duct  which 
can  propagate  waves  with  very  little  attenu- 
ation over  great  distances  in  a manner  similar 
to  the  propagation  of  waves  through  a wave 
guide.  Guided  propagation  through  a duct  in 
the  atmosphere  can  give  quite  remarkable 
transmission  conditions  (figure  15).  However, 
such  ducts  usually  are  formed  only  on  an  over- 
water  path.  The  depth  of  the  duct  over  the 
water’s  surface  may  be  only  20  to  50  feet,  or 
it  may  be  1000  feet  deep  or  more.  Ducts  ex- 
hibit a low-frequency  cutoff  characteristic 
similar  to  a wave  guide.  The  cutoff  frequency 
is  determined  by  depth  of  the  duct  and  by  the 
strength  of  the  discontinuity  in  refractive  in- 
dex at  the  upper  surface  of  the  duct.  The  low- 
est 'frequency  that  can  be  propagated  by  such 
a duct  seldom  goes  below  50  Me.,  and  usually 
will  be  greater  than  100  Me.  even  along  the 
Pacific  Coast. 

Stratospheric  Communication  by  virtue  of 

Reflection  stratospheric  reflection  can  be 

brought  about  during  magnetic 
storms,  aurora  borealis  displays,  and  during 
meteor  showers.  Dx  communication  during  ex- 
tensive meteor  showers  is  characterized  by 
frequent  bursts  of  great  signal  strength  fol- 
lowed by  a rapid  decline  in  strength  of  the 
received  signal.  The  motion  of  the  meteor 
forms  an  ionized  trail  of  considerable  extent 
which  can  bring  about  effective  reflection  of 
signals.  However,  the  ionized  region  persists 
only  for  a matter  of  seconds  so  that  a shower 
of  meteors  is  necessary  before  communication 
becomes  possible. 

The  type  of  communication  which  is  possible 
during  visible  displays  of  the  aurora  borealis 


and  during  magnetic  storms  has  been  called 
aurora-type  dx.  These  conditions  teach  a max- 
imum somewhat  after  the  sunspot  cycle  peak, 
possibly  because  the  spots  on  the  sun  ate 
neater  to  its  equator  (and  more  directly  in  line 
with  the  earth)  in  the  latter  part  of  the  cycle. 
Ionospheric  storms  generally  accompany  mag- 
netic storms.  The  normal  layers  of  the  iono- 
sphere may  be  churned  or  broken  up,  making 
radio  transmission  over  long  distances  diffi- 
cult or  impossible  on  high  frequencies.  Un- 
usual conditions  in  the  ionosphere  sometimes 
modulate  v-h-f  waves  so  that  a definite  tone  or 
noise  modulation  is  noticed  even  on  transmit- 
ters located  only  a few  miles  away. 

A peculatity  of  this  type  of  auroral  propaga- 
tion of  v-h-f  signals  in  the  northern  hemisphere 
is  that  directional  antennas  usually  must  be 
pointed  in  a northerly  direction  for  best  results 
for  transmission  or  reception,  regardless  of 
the  direction  of  the  other  station  being  con- 
tacted. Distances  out  to  700  or  800  miles  have 
been  covered  during  magnetic  storms,  using 
30  and  50  Me.  transmitters,  with  little  evi- 
dence of  any  silent  zone  between  the  stations 
communicating  with  each  other.  Generally, 
voice-modulated  transmissions  are  difficult  or 
impossible  due  to  the  tone  or  noise  modulation 
on  the  signal.  Most  of  the  communication  of 
this  type  has’  taken  place  by  c.w.  or  by  tone 
modulated  waves  with  a keyed  carrier. 


Ionospheric 

Propagation 


Propagation  of  radio  waves  for  communica- 
tion on  frequencies  between  perhaps  3 and 
30  Me.  is  normally  carried  out  by  virtue  of 
ionospheric  reflection  or  refraction.  Under  con- 
ditions of  abnormally  high  ionization  in  the 
ionosphere,  communication  has  been  known 
to  have  taken  place  by  ionospheric  reflection 
on  frequencies  higher  than  50  Me. 

The  ionosphere  consists  of  layers  of  ionized 
gas  located  above  the  stratosphere,  and  ex- 
tending up  to  possibly  300  miles  above  the 
earth.  Thus  we  see  that  high-frequency  radio 
waves  may  travel  over  short  distances  in  a 
direct  line  from  the  transmitter  to  the  receiver, 
or  they  can  be  radiated  upward  into  the  iono- 
sphere to  be  bent  downward  in  an  indirect  ray, 
returning  to  earth  at  considerable  distance 
from  the  transmitter.  The  wave  reaching  a re- 
ceiver via  the  ionosphere  route  is  termed  a 
sky  wave.  The  wave  reaching  a receiver  by 
traveling  in  a direct  line  from  the  transmitting 
antenna  to  the  receiving  antenna  is  commonly 
called  a ground  wave. 

The  amount  of  bending  at  the  ionosphere 


HANDBOOK 


Ionospheric  Propagation  417 


Figure  16 

IONIZATION  DENSITY  IN  THE  IONO- 
SPHERE 

Showing  typical  ionixation  density  of  the 
ionosphere  in  mid-summer.  Note  that  the  Fj 
and  D layers  disappear  at  night,  and  that  the 
density  of  the  E layer  falls  to  such  a low 
value  that  It  is  Ineffective. 


which  the  sky  wave  can  undergo  depends  up- 
on its  frequency,  and  the  amount  of  ionization 
in  the  ionosphere,  which  is  in  turn  dependent 
upon  radiation  from  the  sun.  The  sun  increases 
the  density  of  the  ionosphere  layers  (figure  16) 
and  lowers  their  effective  height.  For  this 
reason,  the  ionosphere  acts  very  differently 
at  different  times  of  day,  and  at  different  times 
of  the  year. 

The  higher  the  frequency  of  a radio  wave, 
the  farther  it  penetrates  the  ionosphere,  and 
the  less  it  tends  to  be  bent  back  toward  the 
earth.  The  lower  the  frequency,  the  mote  easily 
the  waves  are  bent,  and  the  less  they  pene- 
trate the  ionosphere.  160-meter  and  80-meter 
signals  will  usually  be  bent  back  to  etu-th 
even  when  sent  straight  up,  and  may  be  con- 
sidered as  being  reflected  rather  than  refract- 
ed. As  the  frequency  is  raised  beyond  about 
5,000  kc.  (dependent  upon  the  critical  frequen- 
cy of  the  ionosphere  at  the  moment),  it  is 
found  that  waves  transmitted  at  angles  higher 
than  a certain  critical  angle  never  return  to 
earth.  Thus,  on  the  higher  frequencies,  it  is 
necessary  to  confine  radiation  to  low  angles, 
since  the  high  angle  waves  simply  penetrate 
the  ionosphere  and  are  lost. 

The  Fj  Layer  The  higher  of  the  two  major 

reflection  regions  of  the  iono- 
sphere is  called  the  Fj  layer.  This  layer  has 


a virtual  height  of  approximately  175  miles  at 
night,  and  in  the  daytime  it  splits  up  into  two 
layers,  the  upper  one  being  c^led  the  layer 
and  the  lower  being  called  the  F^  layer.  The 
height  of  the  Fj  layer  during  daylight  hours  is 
normally  about  250  miles  on  the  average  and 
the  F,  layer  often  has  a height  of  as  low  as 
140  miles.  It  is  the  F^  layer  which  supports 
all  nighttime  dx  communication  and  nearly  all 
daytime  dx  propagation. 

The  E Layer  Below  the  Fj  layer  is  another 
layer,  called  the  E layer,  which 
is  of  importance  in  daytime  communication 
over  moderate  distances  in  the  frequency  range 
between  3 and  8 Me.  This  layer  has  an  almost 
constant  height  at  about  70  miles.  Since  the 
re-combination  time  of  the  ions  at  this  height 
is  rather  short,  the  E layer  disappears  almost 
completely  a short  time  after  local  sunset. 

The  D Layer  Below  the  E layer  at  a height  of 
about  35  miles  is  an  absorbing 
layer,  called  the  D layer,  which  exists  in  the 
middle  of  the  day  in  the  summertime.  The  layer 
also  exists  during  midday  in  the  winter  time 
during  periods  of  high  solar  activity,  but  the 
layer  disappears  completely  at  night.  It  is  this 
layer  which  causes  high  absorption  of  signals 
in  the  medium  and  high-frequency  range  during 
the  middle  of  the  day. 

Critical  Frequency  The  critical  frequency  of 
an  ionospheric  layer  is  the 
highest  frequency  which  will  be  reflected  when 
the  wave  strikes  the  layer  at  vertical  inci- 
dence. The  critical  frequency  of  the  most  high- 
ly ionized  layer  of  the  ionosphere  may  be  as 
low  as  2 Me.  at  night  and  as  high  as  12  to  13 
Me.  in  the  middle  of  the  day.  The  critical  fre- 
quency is  directly  of  interest  in  that  a skip- 
distance  zone  will  exist  on  all  frequencies 
greater  than  the  highest  critical  frequency  at 
that  time.  The  critical  frequency  is  a measure 
of  the  density  of  ionization  of  the  reflecting 
layers.  The  higher  the  critical  frequency  the 
greater  the  density  of  ionization. 

Maximum  Usable  The  maximum  usable  fre- 
Frequency  quency  or  m.a.f.  is  of  great 

importance  in  long-distance 
communication  since  this  frequency  is  the  high- 
est that  can  be  used  for  communication  be- 
tween any  two  specified  areas.  The  m.u.f.  is 
the  highest  frequency  at  which  a wave  pro- 
jected into  space  in  a certain  direction  will 
be  returned  to  earth  in  a specified  region  by 
ionospheric  reflection.  The  m.u.f.  is  highest 
at  noon  or  in  the  early  afternoon  and  is  high- 
est in  periods  of  greatest  sunspot  activity, 
often  going  to  frequencies  higher  than  50  Me. 
(figure  17). 


418  Radiation,  Propagation  and  Lines 


the  radio 


Figure  17 

TYPICAL  CURVES  SHOWING  CHANGE  IN 
M.U.F.  AT  MAXIMUM  AND  MINIMUM 
POINTS  IN  SUNSPOT  CYCLE 


The  m.u.f.  often  drops  to  frequencies  below 
10  Me.  in  the  early  morning  hours.  The  high 
m.u.f.  in  the  middle  of  the  day  is  brought  about 
by  reflection  from  the  Fj  layer.  M.u.f.  data  is 
published  periodically  in  the  magazines  de- 
voted to  amateur  work,  and  the  m.u.f.  can  be 
calculated  with  the  aid  of  Basic  Radio  Propa- 
gation Predictions,  CRPL-D,  published  month- 
ly by  the  Government  Printing  Office,  Wash- 
ington, D.C. 

Absorption  and  The  optimum  working  fre- 

Optimum  Working  quency  for  any  particular 
Frequency  direction  and  distance  is 

usually  about  15  per  cent 
less  than  the  m.u.f.  for  contact  with  that  par- 
ticular location.  The  absorption  by  the  iono- 
sphere becomes  greater  and  greater  as  the 
operating  frequency  is  progressively  lowered 
below  the  m.u.f.  It  is  this  condition  which 
causes  signals  to  increase  tremendously  in 
strength  on  the  14-Mc.  and  28-Mc.  bands  just 
before  the  signals  drop  completely  out.  At  the 
time  when  the  signals  are  greatest  in  ampli- 
tude the  operating  frequency  is  equal  to  the 
m.u.f.  Then  as  the  signals  drop  out  the  m.u.f. 
has  become  lower  than  the  operating  frequency. 

Skip  Distance  The  shortest  distance  from  a 
transmitring  location  at  which 
signals  reflected  from  the  ionosphere  can  be 
returned  to  the  earth  is  called  the  skip  dis- 
tance. As  was  mentioned  above  under  Critical 
Frequency  there  is  no  skip  distance  for  a fre- 
quency below  the  critical  frequency  of  the 


most  highly  ionized  layer  of  the  ionosphere 
at  the  time  of  transmission.  However,  the  skip 
distance  is  always  present  on  the  14-Mc.  band 
and  is  almost  always  present  on  the  3.5-Mc. 
and  7-Mc.  bands  at  night.  The  actual  measure 
of  the  skip  distance  is  the  distance  between 
the  point  where  the  ground  wave  falls  to  zero 
and  the  point  where  the  sky  wave  begins  to 
return  to  earth.  This  distance  may  vary  from 
40  to  50  miles  on  the  3.5-Mc.  band  to  thou- 
sands of  miles  on  the  28-Mc.  band. 

The  Sporadic- E Occasional  patches  of  ex- 

Layer  tremely  high  ionization  den- 

sity appear  at  intervals 
throughout  the  year  at  a height  approximately 
equal  to  that  of  the  E layer.  These  patches, 
called  the  sporadic-E  layer  may  be  very  small 
or  may  be  up  to  several  hundred  miles  in  ex- 
tent. The  critical  frequency  of  the  sporadic-E 
layer  may  be  greater  than  twice  that  of  the 
normal  ionosphere  layers  which  exist  at  the 
same  time. 

It  is  this  sporadic-E  condition  which  pro- 
vides "short-skip”  contacts  from  400  to  per- 
haps 1200  miles  on  the  28-Mc.  band  in  the 
evening.  It  is  also  the  sporadic-E  condition 
which  provides  the  mote  common  type  of  "band 
opening”  experienced  on  the  50-Mc.  band  when 
very  loud  signals  are  received  from  stations 
from  400  to  1200  miles  distant. 

Cycles  in  The  ionization  density  of 

Ionosphere  Activity  the  ionosphere  is  deter- 
mined by  the  amount  of 
radiation  (probably  ultra  violet)  which  is  be- 
ing received  from  the  sun.  Consequently,  iono- 
sphere activity  is  a function  of  the  amount  of 
radiation  of  the  proper  character  being  emitted 
by  the  sun  and  is  also  a function  of  the  rela- 
tive aspect  of  the  regions  in  the  vicinity  of 
the  location  under  discussion  to  the  sun.  There 
are  four  main  cycles  in  ionosphere  activity. 
These  cycles  are:  the  daily  cycle  which  is 
brought  about  by  the  rotation  of  the  earth,  the 
27-day  cycle  which  is  caused  by  the  rotation 
of  the  sun,  the  seasonal  cycle  which  is  caused 
by  the  movement  of  the  earth  in  its  orbit,  and 
the  11-yeat  cycle  which  is  a cycle  in  sunspot 
activity.  The  effects  of  these  cycles  are  super- 
imposed insofar  as  ionosphere  activity  is  con- 
cerned. Also,  the  cycles  ate  subject  to  short 
term  variations  as  a result  of  magnetic  storms 
and  similar  terrestrial  disturbances. 

The  most  recent  minimum  of  the  11-year 
sunspot  cycle  occured  during  the  winter  of 
1954-1955,  and  we  are  currently  moving  up 
the  slope  of  a new  cycle,  the  maximum  of 
which  will  probably  occur  during  the  year 
1958.  The  current  cycle  is  pictured  in  figure  18. 

Fading  The  lower  the  angle  of  radiation  of 
the  wave,  with  respect  to  the  hoti- 


HANDBOOK 


ll'Year  Sunspot  Cycle 


419 


Figure  18 

THE  YEARLY  TREND  OF  THE  ll-YEAR 
SUNSPOT  CYCLE.  RADIO  CONDITIONS  IN 
GENERAL  WILL  IMPROVE  DURING  1955- 
1958  AS  THE  CYCLE  ADVANCES 


Figure  19 

IONOSPHERE-REFLECTION  WAVE  PATHS 

Showing  typical  lonosp/iere-ref/ecf/on  wave 
paths  during  daylight  hours  when  ionization 
density  is  such  that  frequencies  as  high  as 
28  Me.  will  be  returned  to  earth.  The  dis- 
tance between  ground-wave  range  and  that 
range  where  the  ionosphere-reflected  wave  of 
a specific  frequency  first  will  he  returned  to 
earth  is  called  the  skip  distance. 


zon,  the  farther  away  will  the  wave  return  to 
earth,  and  the  greater  the  skip  distance.  The 
wave  can  be  reflected  back  up  into  the  iono- 
sphere by  the  earth,  and  then  be  reflected  back 
down  again,  causing  a second  skip  distance 
area.  The  drawing  of  figure  19  shows  the  mul- 
tiple reflections  possible.  When  the  receiver 
receives  signals  which  have  traveled  over 
more  than  one  path  between  transmitter  and 
receiver,  the  signal  impulses  will  not  all  arrive 
at  the  same  instant,  as  they  do  not  all  travel 
the  same  distance.  When  two  or  more  signals 
arrive  in  the  same  phase  at  the  receiving  an- 
tenna, the  resulting  signal  in  the  receiver  will 
be  quite  strong.  On  the  other  hand,  if  the  sig- 
nals arrive  180°  out  of  phase,  so  they  tend  to 
cancel  each  other,  the  received  signal  will 
drop— perhaps  to  zero  if  perfect  cancellation 
occurs.  This  explains  why  high-frequency 
signals  are  subject  to  fading. 

Fading  can  be  greatly  reduced  on  the  high 
frequencies  by  using  a transmitting  antenna 
with  sharp  vertical  directivity,  thus  cutting 
down  the  number  of  possible  paths  of  signal 
arrival.  A receiving  antenna  with  similar  char- 
acteristics (sharp  vertical  directivity)  will 
further  reduce  fading.  It  is  desirable,  when 
using  antennas  with  sharp  vertical  directivity, 
to  use  the  lowest  vertical  angle  consistent 
with  good  signal  strength  for  the  frequency 
used. 

Scattered  Scattered  reflections  are  random. 

Reflections  diffused,  substantially  isotropic 

reflections  which  are  partly  re- 


sponsible for  reception  within  the  skip  zone, 
and  for  reception  of  signals  from  directions 
off  the  great  circle  path. 

In  a heavy  fog  or  mist,  it  is  difficult  to  see 
the  road  at  night  because  of  the..bright  glare 
caused  by  scattered  reflection  of  the  head- 
light beam  by  the  minute  droplets.  In  fact,  the 
road  directly  to  the  side  of  the  car  will  be 
weakly  illuminated  under  these  conditions, 
whereas  it  would  not  on  a clear  night  (assum- 
ing flat,  open  country).  This  is  a good  example 
of  propagation  of  waves  by  scattered  reflec- 
tions into  a zone  which  otherwise  would  not 
be  illuminated. 

Scattering  occurs  in  the  ionosphere  at  all 
times,  because  of  irregularities  in  the  medi- 
um (which  result  in  "patches”  corresponding 
to  the  water  droplets)  and  because  of  random- 
phase  radiation  due  to  the  collision  or  recom- 
bination of  free  electrons.  However,  the  nature 
of  the  scattering  varies  widely  with  time,  in 
a random  fashion.  Scattering  is  particularly 
prevalent  in  the  E region,  but  scattered  re- 
flections may  occur  at  any  height,  even  well 
out  beyond  the  virtual  height  of  the  Fj  layer. 

There  is  no  "critical  frequency”  or  "low- 
est perforating  frequency”  involved  in  the 
scattering  mechanism,  though  the  intensity 
of  the  scattered  reflections  due  to  typical 
scattering  in  the  E region  of  the  ionosphere 
decreases  with  frequency. 

When  the  received  signal  is  due  primarily 
to  scattered  reflections,  as  is  the  case  in  the 
skip  zone  or  where  the  great  circle  path  does 
not  provide  a direct  sky  wave  (due  to  low  crit- 
ical or  perforation  frequency,  or  to  an  iono- 
sphere storm)  very  bad  distortion  will  be  evi- 


42  0 Radiation,  Propagation  and  Lines 


THE  RADIO 


dent,  particularly  a "flutter  fade”  and  a char- 

acteristic "hollow”  or  echo  effect. 

Deviations  from  a great  circle  path  ate  es- 
pecially noticeable  in  the  case  of  great  circle 
paths  which  cross  or  pass  near  the  auroral 
zones,  because  in  such  cases  there  often  is 
complete  or  nearly  complete  absorption  of  the 
direct  sky  wave,  leaving  off-path  scattered 
reflections  the  only  mechanism  of  propagation. 
Under  such  conditions  the  predominant  wave 
will  appear  to  arrive  from  a direction  closer 
to  the  equator,  and  the  signal  will  be  notice- 
ably if  not  considerably  weaker  than  a direct 
sky  wave  which  is  received  under  favorable 
conditions. 

Irregular  reflection  of  radio  waves  from 
"scattering  patches”  is  divided  into  two  cate- 
gories: "short  scatter”  and  "long  scatter”. 

Short  scatter  is  the  scattering  that  occurs 
when  a radio  wave  first  reaches  the  scattering 
patches  or  media.  Ordinarily  it  is  of  no  parti- 
cular benefit,  as  in  most  cases  it  only  serves 
to  fill  in  the  inner  portion  of  the  skip  zone 
with  a weak,  distorted  signal. 

Long  scatter  occurs  when  a wave  has  been 
refracted  from  the  F,  layer  and  strikes  scatter- 
ing patches  or  media  on  the  way  down.  When 
the  skip  distance  exceeds  several  hundred 
miles,  long  scatter  is  primarily  responsible 
for  reception  within  the  skip  zone,  particu- 
larly the  outer  portion  of  the  skip  zone.  Dis- 
tortion is  much  less  severe  than  in  the  case 
of  short  scatter,  and  while  the  signal  is  like- 
wise weak,  it  sometimes  can  be  utilized  for 
satisfactory  communication. 

During  a severe  ionosphere  disturbance  in 
the  north  auroral  zone,  it  sometimes  is  possible 
to  maintain  communication  between  the  Eastern 
United  States  and  Northern  Europe  by  the  fol- 
lowing mechanism:  That  portion  of  the  energy 
which  is  radiated  in  the  direction  of  the  great 
circle  path  is  completely  absorbed  upon  reach- 
ing the  auroral  zone.  However,  the  portion  of 
the  wave  leaving  the  United  States  in  a south- 
easterly direction  is  refracted  downward  from 
rhe  Fj  layer  and  encounters  scattering  patches 
or  media  on  its  downward  trip  at  a distance 
of  approximately  2000  miles  from  the  trans- 
mitter. There  it  is  reflected  by  "long  scatter” 
in  all  directions,  this  scattering  region  acting 
like  an  isotropic  radiator  fed  with  a very  small 
fraction  of  the  original  transmitter  power.  The 
great  circle  path  from  this  southerly  point  to 
northern  Europe  does  not  encounter  unfavor- 
able ionosphere  conditions,  and  the  wave  is 
propagated  the  rest  of  the  trip  as  though  it  had 
been  radiated  from  the  scattering  region. 

Another  type  of  scatter  is  produced  when 
a sky  wave  strikes  certain  areas  of  the  earth. 
Upon  striking  a comparatively  smooth  surface 
such  as  the  sea,  there  is  little  scattering,  the 
wave  being  shot  up  again  by  what  could  be 


considered  specular  or  mirror  reflection.  But 

upon  striking  a mountain  range,  for  instance, 
the  reradiation  or  reflected  energy  is  scattered, 
some  of  it  being  directed  back  towards  the 
transmitter,  thus  providing  another  mechanism 
for  producing  a signal  within  the  skip  zone. 

Meteors  and  When  a meteor  strikes  the  earth’s 
"Bursts”  atmosphere,  a cylindrical  region 
of  free  electrons  is  formed  at 
approximately  the  height  of  the  E layer.  This 
slender  ionized  column  is  quite  long,  and  when 
first  formed  is  sufficiently  dense  to  reflect 
radio  waves  back  to  earth  most  readily,  in- 
cluding v-h-f  waves  which  are  not  ordinarily 
returned  by  the  Fj  layer. 

The  effect  of  a single  meteor,  of  normal  size, 
shows  up  as  a sudden  "burst”  of  signal  of 
short  duration  at  points  not  ordinarily  reached 
by  the  transmitter.  After  a period  of  from  10 
to  40  seconds,  recombination  and  diffusion 
have  progressed  to  the  point  where  the  effect 
of  a single  fairly  large  meteor  is  not  percep- 
tible. However,  there  are  many  small  meteors 
impinging  upon  earth’s  atmosphere  every  min- 
ute, and  the  aggregate  effect  of  their  transient 
ionized  trails,  including  the  small  amount  of 
residual  ionization  that  exists  for  several 
minutes  after  the  original  flash  but  is  too  weak 
and  dispersed  to  prolong  a "burst”,  is  be- 
lieved to  contribute  to  the  existence  of  the 
"nighttime  E”  layer,  and  perhaps  also  to 
sporadic  E patches. 

While  there  are  many  of  these  very  small 
meteors  striking  the  earth’s  atmosphere  every 
minute,  meteors  of  normal  size  (sufficiently 
large  to  produce  individual  "bursts”)  do  not 
strike  nearly  so  frequently  except  during  some 
of  the  comparatively  rare  meteor  "showers”. 
During  one  of  these  displays  a "quivering” 
ionized  layer  is  produced  which  is  intense 
enough  to  return  signals  in  the  lower  v-h-f 
range  with  good  strength,  but  with  a type  of 
"flutter”  distortion  which  is  characteristic 
of  this  type  of  propagation. 


21-9  Transmission  Lines 

For  many  reasons  it  is  desirable  to  place 
an  antenna  or  radiating  system  as  high  and  in 
the  clear  as  is  physically  possible,  utilizing 
some  form  of  nonradiating  transmission  line 
to  carry  energy  with  as  little  loss  as  possible 
from  the  transmitter  to  the  radiating  antenna, 
and  conversely  from  the  antenna  to  the  re- 
ceiver. 

There  are  many  different  types  of  transmis- 
sion lines  and,  generally  speaking,  practically 
any  type  of  transmission  line  or  feeder  system 
may  be  used  with  any  type  of  antenna.  How- 


HANDBOOK 


Transmission  Lines  421 


ever,  mechanical  or  electrical  considerations 
often  make  one  type  of  ttansmission  line  better 
adapted  for  use  to  feed  a particular  type  of 
antenna  than  any  other  type. 

Transmission  lines  for  carrying  r-f  energy 
are  of  two  general  types:  non-resonant  and 
resonant,  A non-resonant  transmission  line 
is  one  on  which  a successful  effort  has  been 
made  to  eliminate  reflections  from  the  termi- 
nation (the  antenna  in  the  ttansmitting  case 
and  the  receiver  for  a receiving  antenna)  and 
hence  one  on  which  standing  waves  do  not 
exist  or  are  relatively  small  in  magnitude.  A 
resonant  line,  on  the  other  hand,  is  a trans- 
mission line  on  which  standing  waves  of  ap- 
preciable magnitude  do  appear,  either  through 
inability  to  match  the  characteristic  impedance 
of  the  line  to  the  termination  or  through  in- 
tentional design. 

The  principal  types  of  transmission  line  in 
use  or  available  at  this  time  include  the  open- 
wire  line  (two-wire  and  four-wire  types),  two- 
wite  solid-dielectric  line  ("Twin-Lead”  and 
similar  ribbon  or  tubular  types),  two-wire  poly- 
ethylene-filled shielded  line,  coaxial  line  of 
the  solid-dielectric,  beaded,  stub-supported, 
or  pressurized  type,  rectangular  and  cylindrical 
wave  guide,  and  the  single-wire  feeder  oper- 
ated against  ground.  The  significant  charac- 
teristics of  the  more  popular  types  of  trans- 
mission line  available  at  this  time  are  given 
in  the  chart  of  figure  21. 


21-10  Non-Resonant 

Transmission  Lines 

A non-resonant  or  untuned  transmission  line 
is  a line  with  negligible  standing  waves. 
Hence,  a non-resonant  line  is  a line  carrying 
r-f  power  only  in  one  direction— from  the  source 
of  energy  to  the  load. 

Physically,  the  line  itself  should  be  iden- 
tical throughout  its  length.  There  will  be  a 
smooth  distribution  of  voltage  and  current 
throughout  its  length,  both  tapering  off  very 
slightly  towards  the  load  end  of  the  line  as  a 
result  of  line  losses.  The  attenuation  (loss) 
in  cettain  types  of  untuned  lines  can  be  kept 
very  low  for  line  lengths  up  to  several  thou- 
sand feet.  In  other  types,  patticularly  where 
the  dielectric  is  not  air  (such  as  in  the  twisted- 
pair line),  the  losses  may  become  excessive 
at  the  higher  frequencies,  unless  the  line  is 
relatively  short. 

Tronsmission-Line  All  transmission  lines  have 
Impedance  distributed  inductance, 

capacitance  and  resist- 
ance. Neglecting  the  resistance,  as  it  is  of 
minor  importance  in  short  lines,  it  is  found 


Figure  20 

CHARACTERISTIC  IMPEDANCE  OF  TYPI- 
CAL TWO-WIRE  OPEN  LINES 


that  the  inductance  and  capacitance  per  unit 
length  determine  the  characteristic  or  surge 
impedance  of  the  line.  Thus,  the  surge  im- 
pedance depends  upon  the  nature  and  spacing 
of  the  conductors,  and  the  dielectric  sepa- 
rating them. 

Speaking  in  electrical  terms,  the  charac- 
teristic impedance  of  a transmission  line  is 
simply  the  ratio  of  the  voltage  across  the  line 
to  the  current  which  is  flowing,  the  same  as 
is  the  case  with  a simple  resistor:  = E/l. 

Also,  in  a substantially  loss-less  line  (one 
whose  attenuation  per  wavelength  is  small) 
the  energy  stored  in  the  line  will  be  equally 
divided  between  the  capacitive  field  and  the 
inductive  field  which  serve  to  propagate  the 
energy  along  the  line.  Hence  the  character- 
istic impedance  of  a line  maybe  expressed  as: 

z„  = VlTc. 

Two-Wire  A two-wire  transmission  system 
Open  Line  is  easy  to  construct.  Its  surge  im- 
pedance can  be  calculated  quite 
easily,  and  when  properly  adjusted  and  bal- 
anced to  ground,  with  a conductor  spacing 
which  is  negligible  in  terms  of  the  wave- 
length of  the  signal  carried,  undesirable  feeder 
radiation  is  minimized;  the  curtent  flow  in 
the  adjacent  wires  is  in  opposite  directions, 
and  the  magnetic  fields  of  the  two  wires  are 
in  opposition  to  each  other.  When  a two-wire 
line  is  terminated  with  the  equivalent  of  a 
pure  resistance  equal  to  the  characteristic 
impedance  of  the  line,  the  line  becomes  a non- 
resonant  line. 

Expressed  in  physical  terms,  the  character- 
istic impedance  of  a two-wire  open  line  is 
equal  to: 


422  Radiation,  Propagation  and  Lines 


THE  RADIO 


CHARACTERISTICS  OF  COMMON  TRANSMISSION  LINES 


ATTENUATION 

db/100  FEET 
VSWR=  1.0 

VELO- 

CITY 

gj 

REMARKS 

-ACTOR 

V 

OPEN  WIRELINE,  N«t2 
COPPER. 

B 

B 

0.96-0.99 

- 

BASED  UPON  4' SPACING  BELOW  50  MC.i  2' SPACING  ABOVE  50  MC.  RADIATION 
LOSSES  INCLUDED.  CLEAN.  LOW  LOSS  CERAMIC  INSULATION  ASSUMED.  RADIATION 
HIGH  ABOVE  150  MC. 

RIBBON  LINE,  REC.TYPE, 
300  OHMS. 

(7/26  CONDUCTORS) 

0.66 

2.2 

B 

0.82 

FOR  CLEAN.  DRY  LINE.  WET  WEATHER  PERFORMANCE  RATHER  POOR.  BEST  LINE  IS 
SLIGHTLY  CONVEX.  AVOID  LINE  THAT  HAS  CONCAVE  DIELECTRIC.  SUITABLE  FOR 
LOW  POWER  TRANSMITTING  APPLICATIONS.  LOSSES  INCREASE  AS  LINE  WEATHERS. 
HANDLES  400  WATTS  AT  30  MC.  IF  VSWR  IS  LOW. 

TUBULAR  "TWIN-LEAD" 
REC.  TYPE.  300  OHMS, 
3/16' O.D..  (AMPHENOL 
TYPE  14-271) 

■ 

0 

B 

- 

- 

CHARACTERISTICS  SIMILAR  TO  RECEIVING  TYPE  RIBBON  LINE  EXCEPT  FOR  MUCH 
BETTER  WET  WEATHER  PERFORMANCE. 

B 

B 

- 

- 

- 

CHARACTERISTICS  VARY  SOMEWHAT  WITH  MANUFACTURER,  BUT  APPROXIMATE 
THOSE  OF  RECEIVING  TYPE  RIBBON  EXCEPT  FOR  GREATER  POWER  HANDLING 
CAPABILITY  Aftf)  SLIGHTLY  BETTER  WET  WEATHER  PERFORMANCE. 

iisii 

Q 

B 

B 

0.79 

B 

FOR  USE  WHERE  RECEIVING  TYPE  TUBULAR  'TWIN-LEAD'  DOES  NOT  HAVE  SUFFI- 
CIENT POWER  HANDLING  CAPABILITY.  WILL  HANDLE  1 KW  AT  30  MC.  IF  VSWR 

IS  LOW. 

RIBBON  LINE,  RECEIVE. 
TYPE,  130  OHMS. 

D 

Q 

m 

0.77'^ 

m 

USEFUL  FOR  QUARTER  WAVE  MATCHING  SECTIONS.  NO  LONGER  WIDELY  USED 

AS  A LINE. 

RIBBON  UNE,  RECEIVE. 
TYPE,  75  OHMS. 

2.0 

5.0 

o.eo'*' 

BB 

USEFUL  MAINLY  IN  THE  H-F  RANGE  BECAUSE  OF  EXCESSIVE  LOSSES  AT  V-H-F 
AND  U-H-F.  LESS  AFFECTED  BY  WEATHER  THAN  300  OHM.RIBBON. 

RIBBON  LINE.  TRANS. 
TYPE,  75  OHMS. 

1.5 

3.9* 

6.0 

0,7.» 

BSI 

VERY  SATISFACTORY  FOR  TRANSMITTING  APPLICATIONS  BELOW  30  MC.  AT 
POWERS  UP  TO  1 KW.  NOT  SIGNIFICANTLY  AFFECTED  BY  WET  WEATHER. 

RG-e/U  COAX  (520HM3) 

1.0 

wn 

KB 

WILL  handle  2KW  AT  30  MC.  IF  VSWR  IS  LOW.  0.4' O.D.  7/21  CONDUCTOR. 

RG-11/UCOAX(75  OHMS) 

0.94 

■Q 

KB 

ma 

WILL  HANDLE  1.4  KW  AT  30  MC.  IF  VSWR  IS  LOW.  0.4"O.D.  7/26  CONDUCTOR . 

R6-17/U  COAX  (52  OHMS) 

IKI 

WILL  HANDLE  7.6  KW.  AT  30  MC.  IF  VSWR  IS  LOW.  0.67"  O.D.  0.19"  DIA.  CONDUCTOR 

RG-38/U  COAX  (53  OHM^ 

Bl 

Bl 

iQnm 

WILL  HANDLE  430  WATTS  AT  30  MC.  IF  VSWR  IS  LOW.  0-2"  QO.  N*  20  CONDUCTOR. 

RG-59/U  COAX  (73  OHMS) 

m 

m 

Bl 

BB 

WILL  HANDLE  660  WATTS  AT  30  MC.  IF  VSWR  IS  LOW.  0 24"0.0.  N«22  CONDUCTOR. 

TV-59  COAX  (72  OHMS) 

ra 

g 

B 

0.66 

22 

'COMMERCIAL'  VERSION  OF  R6-50/U  FOR  LESS  EXACTING  APPLICATIONS.  LESS 
EXPENSIVE. 

RG-22/U  SHIELDED 
PAIR  (95  OHMS) 

1.7 

ID 

B 

0.66 

16 

'FOR  SHIELDED,  BALANCED-TO-GROUND  APPLICATIONS.  VERY  LOW  NOISE 
PICKUP.  0.4"0.0. 

IB 

ID 

ID 

Hi 

4 

DESIGNED  FOR  TV  LEAD-IN  IN  NOISY  LOCATIONS.  LOSSES  HIGHER  THAN 
REGULAR  300  OHM  RIBBON,  BUT  DO  NOT  INCREASE  AS  MUCH  FROM  WEATHERING. 

'if  APPROXIMATE.  EXACT  FIGURE  VARIES  SLIGHTLY  WITH  MANUFACTURER. 


FIGURE  21 


2S 

Zo  = 276  log^ - 
d 

Where: 

S is  the  exact  distance  between  wite  centers 
in  some  convenient  unit  of  measurement,  and 
d is  the  diameter  of  the  wite  measured  in  the 
same  units  as  the  wire  spacing,  S. 

2S 

Since  expresses  a ratio  only,  the  units 

d 

of  measurement  may  be  centimeters,  milli- 
meters, or  inches.  This  makes  no  difference 
in  the  answer,  so  long  as  the  substituted 
values  for  S and  d are  in  the  same  units. 

The  equation  is  accurate  so  long  as  the 
wire  spacing  is  relatively  large  as  compared 
to  the  wire  diameter. 


Surge  impedance  values  of  less  than  200 
ohms  ate  seldom  used  in  the  open-type  two- 
wire  line,  and,  even  at  this  rather  high  value 
of  Zo  the  wite  spacing  S is  uncomfortably 
close,  being  only  5.3  times  the  wire  diameter  d. 

Figure  20  gives  in  graphical  form  the  surge 
impedance  of  practicable  two-wire  lines.  The 
chart  is  self-explanatory,  and  is  sufficiently 
accurate  for  practical  purposes. 

Ribbon  and  Instead  of  using  spacer  in- 

Tubular  Trans-  sulators  placed  periodically 

mission  Line  along  the  transmission  line 

it  is  possible  to  mold  the 
line  conductors  into  a ribbon  or  tube  of  flex- 
ible low-loss  dielectric  material.  Such  line, 
with  polyethylene  dielectric,  is  used  in  enor- 
mous quantities  as  the  lead-in  transmission 
line  for  FM  and  TV  receivers.  The  line  is 
available  from  several  manufacturers  in  the 


THE  RADIO 


Transmission  Lines  423 


ribbon  and  tubular  configuration,  with  char- 
acteristic impedance  values  from  75  to  300 
ohms.  Receiving  types,  and  transmitting  types 
for  power  levels  up  to  one  kilowatt  in  the  h-f 
range,  are  listed  with  their  pertinent  char- 
acteristics, in  the  table  of  figure  21. 

Coaxial  Line  Several  types  of  coaxial  cable 
have  come  into  wide  use  for 
feeding  power  to  an  antenna  system.  A cross- 
sectional  view  of  a coaxial  cable  (sometimes 
called  concentric  cable  or  line)  is  shown  in 
figure  22. 

As  in  the  parallel-wire  line,  the  power  lost 
in  a properly  terminated  coaxial  line  is  the 
sum  of  the  effective  resistance  losses  along 
the  length  of  the  cable  and  the  dielectric 
losses  between  the  two  conductors. 

Of  the  two  losses,  the  effective  resistance 
loss  is  the  greater;  since  it  is  largely  due  to 
the  skin  effect,  the  line  loss  (all  other  condi- 
tions the  same)  will  increase  directly  as  the 
square  root  of  the  frequency. 

Figure  22  shows  that,  instead  of  having  two 
conductors  running  side  by  side,  one  of  the 
conductors  is  placed  inside  of  the  other.  Since 
the  outside  conductor  completely  shields  the 
inner  one,  no  radiation  takes  place.  The  con- 
ductors may  both  be  tubes,  one  within  the 
other;  the  line  may  consist  of  a solid  wire 
within  a tube,  or  it  may  consist  of  a stranded 
or  solid  inner  conductor  with  the  outer  con- 
ductor made  up  of  one  or  two  wraps  of  copper 
shielding  braid. 

In  the  type  of  cable  most  popular  for  mili- 
tary and  non-commercial  use  the  inner  con- 
ductor consists  of  a heavy  stranded  wire,  the 
outer  conductor  consists  of  a braid  of  copper 
wire,  and  the  inner  conductor  is  supported 
within  the  outer  by  means  of  a semi-solid 
dielectric  of  exceedingly  low  loss  character- 
istics called  polyethylene.  The  Army-Navy 
designation  on  one  size  of  this  cable  suitable 
for  power  levels  up  to  one  kilowatt  at  fre- 
quencies as  high  as  30  Me.  is  AN/RG-8/U. 
The  outside  diameter  of  this  type  of  cable  is 
approximately  one-half  inch.  The  character- 
istic impedance  of  this  cable  type  is  52  ohms, 
but  other  similar  types  of  greater  and  smaller 
power-handling  capacity  are  available  in  im- 
pedances of  52,  75,  and  95  ohms. 

When  using  solid  dielectric  coaxial  cable 
it  is  necessary  that  precautions  be  taken  to 
insure  that  moisture  cannot  enter  the  line.  If 
the  better  grade  of  connectors  manufactured 
for  the  line  are  employed  as  terminations,  this 
condition  is  automatically  satisfied.  If  con- 
nectors are  not  used,  it  is  necessary  that 
some  type  of  moisture-proof  sealing  compound 
be  applied  to  the  end  of  the  cable  where  it 
will  be  exposed  to  the  weather. 

Nearby  metallic  objects  cause  no  loss,  and 
coaxial  cable  may  be  run  up  air  ducts  or  ele- 


O 


ZO=  t38  LOG,o 


COAXIAL  OR 


CONCENTRIC  LINE 


D(=insioe  diameter  of 
OUTER  CONDUCTOR 


D=outsioe  diameter  of 

INNER  CONDUCTOR 


Figure  22 

CHARACTERISTIC  IMPEDANCE  OF  AIR- 
FILLED  COAXIAL  LINES 

If  the  filling  of  the  line  is  a dielectric  mo* 
terial  other  than  air,  the  characteristic  im- 
pedance of  the  line  will  be  reduced  by  a 
factor  proportional  to  the  square-root  of  the 
dielectric  constant  of  the  material  used  as  a 
dielectric  within  the  line. 


vator  shafts,  inside  walls,  or  through  metal 
conduit.  Insulation  troubles  can  be  forgotten. 
The  coaxial  cable  may  be  buried  in  the  ground 
or  suspended  above  ground. 

Standing  Woves  Standing  waves  on  a trans- 
mission line  always  are  the 
result  of  the  reflection  of  energy.  The  only 
significant  reflection  which  takes  place  in  a 
normal  installation  is  that  at  the  load  end  of 
the  line.  But  reflection  can  take  place  from 
discontinuities  in  the  line,  such  as  caused  by 
insulators,  bends,  or  metallic  objects  adjacent 
to  an  unshielded  line. 

When  a uniform  transmission  line  is  termi- 
nated in  an  impedance  equal  to  its  surge  im- 
pedance, reflection  of  energy  does  not  occur, 
and  no  standing  waves  are  present.  When  the 
load  termination  is  exactly  the  same  as  the 
line  impedance,  it  simply  means  that  the  load 
takes  energy  from  the  line  just  as  fast  as  the 
line  delivers  it,  no  slower  and  no  faster. 

Thus,  for  proper  operation  of  an  untuned 
line  (with  standing  waves  eliminated),  some 
form  of  impedance-matching  arrangement  must 
be  used  between  the  transmission  line  and 
the  antenna,  so  that  the  radiation  resistance 
of  the  antenna  is  reflected  back  into  the  line 
as  a nonreactive  impedance  equal  to  the  line 
impedance. 

The  termination  at  the  antenna  end  is  the 
only  critical  characteristic  about  the  untuned 


424  Radiation,  Propagation  and  Lines 


THE  RADIO 


line  fed  by  a transmitter.  It  is  the  reflection 
from  the  antenna  end  which  starts  waves  mov- 
ing back  toward  the  transmitter  end.  When 
waves  moving  in  both  directions  along  a con- 
ductor meet,  standing  waves  are  set  up. 

Semi-Resonant  A well-constructed  open- 

Parollel-Wire  Lines  wire  line  has  acceptably 

low  losses  when  its  length 
is  less  than  about  two  wavelengths  even  when 
the  voltage  standing-wave  ratio  is  as  high  as 
10  to  1.  A transmission  line  constructed  of 
ribbon  or  tubular  line,  however,  should  have 
the  standing-wave  ratio  kept  down  to  not  more 
than  about  3 to  1 both  to  reduce  power  loss 
and  because  the  energy  dissipation  on  the  line 
will  be  localized,  causing  overheating  of  the 
line  at  the  points  of  maximum  current. 

Because  moderate  standing  waves  can  be 
toleratedon  open-wire  lines  without  much  loss, 
a standing- wave  ratio  of  2/1  or  3/1  is  con- 
sidered acceptable  with  this  type  of  line,  even 
when  used  in  an  untuned  system.  Strictly 
speaking,  a line  is  untuned,  or  non-resonant, 
only  when  it  is  perfectly  flat,  with  a standing- 
wave  ratio  of  1 (no  standing  waves).  However, 
some  mismatch  can  be  tolerated  with  open-wire 
untuned  lines,  so  long  as  the  reactance  is  not 
objectionable,  or  is  eliminated  by  cutting  the 
line  to  approximately  resonant  length. 


21-11  Tuned  or 

Resonant  Lines 

If  a transmission  line  is  terminated  in  its 
characteristic  surge  impedance,  there  will  be 
no  reflection  at  the  end  of  the  line,  and  the 
current  and  voltage  distribution  will  be  uni- 
form along  the  line.  If  the  end  of  the  line  is 
either  open-circuited  or  short-circuited,  the 
reflection  at  the  end  of  the  line  will  be  100 
per  cent,  and  standing  waves  of  very  great  am- 
plitude will  appear  on  the  line.  There  will  still 
be  practically  no  radiation  from  the  line  if  it  is 
closely  spaced,  but  voltage  nodes  will  be 
found  every  half  wavelength,  the  voltage  loops 
corresponding  to  current  nodes  (figure  23). 

If  the  line  is  terminated  in  some  value  of 
resistance  other  than  the  characteristic  surge 
impedance,  there  will  be  some  reflection,  the 
amount  being  determined  by  the  amount  of  mis- 
match. With  reflection,  there  will  be  standing 
waves  (excursions  of  current  and  voltage) 
along  the  line,  though  not  to  the  same  extent 
as  with  an  open-circuited  or  short-circuited 
line.  The  current  and  voltage  loops  will  occur 
at  the  same  points  along  the  line  as  with  the 
open  or  short-circuited  line,  and  as  the  ter- 
minating impedance  is  made  to  approach  the 
characteristic  impedance  of  the  line,  the  cur- 


SWR=t.o  Zl=Zo 


SWR  - f.5  Zl=  I.S  or  0.67  Zo 


SWR  *3.0  Zl  * 3.0  OR  0.33  ZO 


SWRs«0  Zu=00R«O 


Figure  23 

STANDING  WAVES  ON  A TRANS- 
MISSION LINE 

As  shown  at  (A),  tha  voltage  and  current  ore 
consfont  on  o transmission  line  which  is 
terminated  in  its  characteristic  impedance, 
assuming  that  losses  are  small  enough  so 
that  they  may  be  neglected.  (B)  shows  the 
variation  in  current  or  in  voltage  on  a line 
terminated  in  a load  with  a reflection  co- 
^fficient  of  0,2  so  that  a standing  wave  ratio 
of  1.5  to  1 is  set  up.  At  (C)  the  reflection 
coefficient  has  been  increased  to  0.5,  with 
the  formation  of  a 3 to  1 standing^wave  ratio 
on  the  line.  At  (D)  the  line  has  been  termi- 
nated in  a load  which  has  a reflection  co- 
efficient of  1.0  (short,  open  circuit,  or  a pure 
reactance)  so  that  all  the  energy  is  reflected 
with  the  formation  of  an  infinite  standing- 
wave  ratio. 


rent  and  voltage  along  the  line  will  become 
more  uniform.  The  foregoing  assumes,  of 
course,  a purely  resistive  (non-reactive)  load. 
If  the  load  is  reactive,  standing  waves  also 
will  be  formed.  But  with  a reactive  load  the 
nodes  will  occur  at  different  locations  from 
the  node  locations  encountered  with  wrong- 
value  resistive  termination. 

A well  built  500-  to  600-ohm  transmission 
line  may  be  used  as  a resonant  feeder  for 
lengths  up  to  several  hundred  feet  with  very 
low  loss,  so  long  as  the  amplitude  of  the 
standing  waves  (ratio  of  maximum  to  minimum 
voltage  along  the  line)  is  not  too  great.  The 


HANDBOOK 


Tuned  Lines  425 


amplitude,  in  turn,  depends  upon  the  mismatch 
at  the  line  termination.  A line  of  no.  12  wire, 
spaced  6 inches  with  good  ceramic  or  plastic 
spreaders,  has  a surge  impedance  of  approx- 
imately 600  ohms,  and  makes  an  excellent 
tuned  feeder  for  feeding  anything  between  60 
and  6000  ohms  (at  frequencies  below  30  Me.). 
If  used  to  feed  a load  of  higher  or  lower  imped- 
ance than  this,  the  standing  waves  become 
great  enough  in  amplitude  that  some  loss  will 
occur  unless  the  feeder  is  kept  short.  At  fre- 
quencies above  30  Me.,  the  spacing  becomes 
an  appreciable  fraction  of  a wavelength,  and 
radiation  from  the  line  no  longer  is  negligible. 
Hence,  coaxial  line  or  close-spaced  parallel- 
wire  line  is  recommended  for  v-h-f  work. 

If  a transmission  line  is  not  perfectly  match- 
ed, it  should  be  made  resonant,  even  though 
the  amplitude  of  the  standing  waves  (voltage 
variation)  is  not  particularly  great.  This  pre- 
vents reactance  from  being  coupled  into  the 
final  amplifier.  A feed  system  having  moderate 
standing  waves  may  be  made  to  present  a non- 
reactive load  to  the  amplifier  either  by  tuning 
or  by  pruning  the  feeders  to  approximate  reso- 
nance. 

Usually  it  is  preferable  with  tuned  feeders 
to  have  a current  loop  (voltage  minimum)  at  the 
transmitter  end  of  the  line.  This  means  that 
when  voltage-  feeding  an  antenna,  the  tuned 
feeders  should  be  made  an  odd  number  of  quar- 
ter wavelengths  long,  and  when  current-feeding 
an  antenna,  the  feeders  should  be  made  an 
even  number  of  quarter  wavelengths  long.  Actu- 
ally, the  feeders  ate  made  about  10  per  cent  of 
a quartet  wave  longer  than  the  calculated 
value  (the  value  given  in  the  tables)  when 
they  are  to  be  series  tuned  to  tesonance  by 
means  of  a capacitor,  instead  of  being  trimmed 
and  pruned  to  resonance. 

When  tuned  feeders  ate  used  to  feed  an  an- 
tenna on  more  than  one  band,  it  is  necessary 
to  compromise  and  make  provision  for  both 
series  and  parallel  tuning,  inasmuch  as  it  is 
impossible  to  cut  a feeder  to  a length  that 
will  be  optimum  for  several  bands.  If  a voltage 
loop  appears  at  the  transmitter  end  of  the  line 
on  certain  bands,  parallel  tuning  of  the  feed- 
ers will  be  required  in  order  to  get  a transfer 
of  energy.  It  is  impossible  to  transfer  energy 
by  inductive  coupling  unless  current  is  flow- 
ing. This  is  effected  at  a voltage  loop  by  the 


presence  of  the  resonant  tank  circuit  formed 
by  parallel  tuning  of  the  antenna'  coil. 


21-12  Line  Discontinuities 

In  the  previous  discussion  we  have  assumed 
a transmission  line  which  was  uniform  through- 
out its  length.  In  actual  practice,  this  is 
usually  not  the  case. 

Whenever  there  is  any  sudden  change  in  the 
characteristic  impedance  of  the  line,  partial 
reflection  will  occur  at  the  point  of  discon- 
tinuity. Some  of  the  energy  will  be  transmitted 
and  some  reflected,  which  is  essentially  the 
same  as  having  some  of  the  energy  absorbed 
and  some  reflected  in  so  far  as  the  effect  upon 
the  line  from  the  generator  to  that  point  is 
concerned.  The  discontinuity  can  by  ascribed 
a reflection  coefficient  just  as  in  the  case  of 
an  unmatched  load. 

In  a simple  case,  such  as  a finite  length  of 
uniform  line  having  a characteristic  impedance 
of  500  ohms  feeding  into  an  infinite  length  of 
uniform  line  having  a characteristic  impedance 
of  100  ohms,  the  behavior  is  easily  predicted. 
The  infinite  100  ohm  line  will  have  no  standing 
waves  and  will  accept  the  same  power  from  the 
500  ohm  line  as  would  a 100  ohm  resistor, 
and  the  rest  of  the  energy  will  be  reflected  at 
the  discontinuity  to  produce  standing  waves 
from  there  back  to  the  generator.  However,  in 
the  case  of  a complex  discontinuity  placed  at 
an  odd  distance  down  a line  terminated  in  a 
complex  impedance,  the  picture  becomes  com- 
plicated, especially  when  the  discontinuity  is 
neither  sudden  nor  gradual,  but  intermediate 
between  the  two.  This  is  the  usual  case  with 
amateur  lines  that  must  be  erected  around 
buildings  and  trees. 

In  any  case,  when  a discontinuity  exists 
somewhere  on  a line  and  is  not  a smooth, 
gradual  change  embracing  several  wavelengths, 
it  is  not  possible  to  avoid  standing  waves 
throughout  the  entire  length  of  the  line.  If  the 
discontinuity  is  sharp  enough  and  is  great 
enough  to  be  significant,  standing  waves  must 
exist  on  one  side  of  the  discontinuity,  and 
may  exist  on  both  sides  in  many  cases. 


CHAPTER  TWENTY-TWO 


Antennas  and  Antenna  Matching 


Antennas  for  the  lower  frequency  portion  of 
the  h-f  spectrum  (perhaps  from  1.8  to  7.0  Me.), 
and  temporary  or  limited  use  antennas  for  the 
upper  portion  of  the  h-f  range,  usually  are  of 
a relatively  simple  type  in  which  directivity 
is  not  a prime  consideration.  Also,  it  often  is 
desirable,  in  amateur  work,  that  a single  an- 
tenna system  be  capable  of  operation  at  least 
on  the  3.5-Mc.  and  7.0-Mc.  range,  and  prefer- 
ably on  other  frequency  ranges.  Consequently, 
the  first  portion  of  this  chapter  will  be  de- 
voted to  a discussion  of  such  antenna  sys- 
tems. The  latter  portion  of  the  chapter  is  de- 
voted to  the  general  problem  of  matching  the 
antenna  transmission  line  to  antenna  systems 
of  the  fixed  type.  Matching  the  antenna  trans- 
mission line  to  the  rotatable  directive  array 
is  discussed  in  Chapter  Twenty-five. 


22-1  End-Fed  Half-Wave 

Horizontal  Antennas 

The  half-wave  horizontal  dipole  is  the  most 
common  and  the  most  practical  antenna  for  the 
3.5-Mc.  and  7-Mc.  amateur  bands.  The  form  of 
the  dipole,  and  the  manner  in  which  it  is  fed 
are  capable  of  a large  number  of  variations. 
Figure  2 shows  a number  of  practicable  forms 
of  the  simple  dipole  antenna  along  with  meth- 
ods of  feed. 


Usually  a high-frequency  doublet  is  mounted 
as  high  and  as  much  in  the  clear  as  possible, 
for  obvious  reasons.  However,  it  is  sometimes 
justifiable  to  bring  part  of  the  radiating  sys- 
tem directly  to  the  transmitter,  feeding  the  an- 
tenna without  benefit  of  a transmission  line. 
This  is  permissible  when  (1)  there  is  insuffi- 
cient room  to  erect  a 75-  or  80-meter  horizon- 
tal dipole  and  feed  line,  (2)  when  a long  wire 
is  also  to  be  operated  on  one  of  the  higher 
frequency  bands  on  a harmonic.  In  either  case, 
it  is  usually  possible  to  get  the  main  portion 
of  the  antenna  in  the  clear  because  of  its 
length.  This  means  that  the  power  lost  by 
bringing  the  anteima  directly  to  the  transmitter 
is  relatively  small. 

Even  so,  it  is  not  best  practice  to  bring  the 
high-voltage  end  of  an  antenna  into  the  oper- 
ating room  because  of  the  increased  difficulty 
in  eliminating  BCI  and  TVI.  For  this  reason 
one  should  dispense  with  a feed  line  in  con- 
junction with  a Hertz  antenna  only  as  a last 
resort. 

End-Fed  The  end-fed  antenna  has  no  form 

Antennas  of  transmission  line  to  couple  it 

to  the  transmitter,  but  brings  the 
radiating  portion  of  the  antenna  right  down  to 
the  transmitter,  where  some  form  of  coupling 
system  is  used  to  transfer  energy  to  the  an- 
tenna. 

Figure  1 shows  two  common  methods  of 
feeding  the  Fuchs  antenna  or  end- fed  Hertz. 


426 


Center-Fed  Antennas  427 


FROM  TRANSMITTER 


Figure  1 

THE  END- FED  HERTZ  ANTENNA 

Showingthe  manner  in  which  an  end'fed  Hertz 
antenna  may  be  fed  through  a tow-impedance 
line  and  low-pass  filter  by  using  a resonant 
tank  circuit  as  at  (A),  or  through  the  use  of 
a reverse-connected  pi  network  as  at  (B), 


Some  harmonic-attenuating  provision  (in  addi- 
tion to  the  usual  low-pass  TVI  filter)  must  be 
included  in  the  coupling  system,  as  an  end- 
fed  antenna  itself  offers  no  discrimination 
against  harmonics,  either  odd  or  even. 

The  end-fed  Hertz  antenna  has  rather  high 
losses  unless  at  least  three-quarters  of  the 
radiator  can  be  placed  outside  the  operating 
room  and  in  the  clear.  As  there  is  r-f  voltage 
at  the  point  where  the  antenna  enters  the 
operating  room,  the  insulation  at  that  point 
should  be  several  times  as  effective  as  the 
insulation  commonly  used  with  low-voltage 
feeder  systems.  This  antetuia  can  be  operated 
on  all  of  its  higher  harmonics  with  good  effi- 
ciency, and  can  be  operated  at  half  frequency 
against  ground  as  a quarter-wave  Marconi. 

As  the  frequency  of  an  antenna  is  raised 
slightly  when  it  is  bent  anywhere  except  at  a 
voltage  or  current  loop,  an  end-fed  Hertz  an- 
tenna usually  is  a few  pet  cent  longer  than  a 
straight  half-wave  doublet  for  the  same  fre- 
quency, because,  ordinarily,  it  is  impractical 
to  bring  a wire  in  to  the  transmitter  without 
making  several  bends. 

The  Zepp  Antenno  The  Zeppelin  or  zepp  an- 
System  tetma  system,  illustrated 

in  figure  2A  is  very  con- 
venient when  it  is  desired  to  operate  a single 
radiating  wire  on  a number  of  harmonically  re- 
lated frequencies. 

The  zepp  antenna  system  is  easy  to  tune, 
and  can  be  used  on  several  bands  by  merely 


retuning  the  feeders.  The  overall  efficiency  of 
the  zepp  antenna  system  is  not  quite  as  high 
for  long  feeder  lengths  as  for  some  of  the  an- 
tenna systems  which  employ  non-resonant 
transmission  lines,  but  where  space  is  limited 
and  where  operation  on  more  than  one  band 
is  desired,  the  zepp  has  some  decided  ad- 
vantages. 

As  the  radiating  portion  of  the  zepp  antenna 
system  must  always  be  some  multiple  of  a 
half  wave  long,  there  is  always  high  voltage 
present  at  the  point  where  the  live  zepp  feed- 
er attaches  to  the  end  of  the  radiating  portion 
of  the  anterma.  Thus,  this  type  of  zepp  an- 
tenna system  is  voltage  jeri. 

Stub-Fed  Zepp-  Figure  2C  shows  a modifica- 
Type  Radiator  tion  of  the  zepp-type  antenna 
system  to  allow  the  use  of 
a non-resonant  transmission  line  between  the 
radiating  portion  of  the  antenna  and  the  trans- 
mitter. The  zepp  portion  of  the  antenna  is 
resonated  as  a quartet-wave  stub  and  the  non- 
resonant  feeders  are  connected  to  the  stub  at 
a point  where  standing  waves  on  the  feeder 
ate  minimized.  The  procedure  for  making  these 
adjustments  is  described  in  detail  in  Section 
22-8  This  type  of  antenna  system  is  quite 
satisfactory  when  it  is  necessary  physically 
to  end  feed  the  antenna,  but  where  it  is  neces- 
sary also  to  use  non-resonant  feeder  between 
the  transmitter  and  the  radiating  system. 


22-2  Center-Fed  Half- 

Wave  Horizontal  Antennas 

The  center  feeding  of  a half-wave  antenna 
system  is  usually  to  be  desired  over  an  end- 
fed  system  since  the  center-fed  system  is  in- 
herently balanced  to  ground  and  is  therefore 
less  likely  to  be  troubled  by  feeder  radiation. 
A number  of  center-fed  systems  ate  illustrated 
in  figure  2. 

The  Tuned  The  current- fed  doublet  with 

Doublet  spaced  feeders,  sometimes 

called  a center- fed  zepp,  is  an 
inherently  balanced  system  if  the  two  legs  of 
the  radiator  ate  electrically  equal.  This  fact 
holds  true  regardless  of  the  frequency,  or  of 
the  harmonic,  on  which  the  system  is  oper- 
ated. The  system  can  successfully  be  oper- 
ated over  a wide  range  of  frequencies  if  the 
system  as  a whole  (both  tuned  feeders  and  the 
center-fed  flat  top)  can  be  resonated  to  the 
operating  frequency.  It  is  usually  possible  to 
tune  such  an  antenna  system  to  resonance 
with  the  aid  of  a tapped  coil  and  a tuning  ca- 
pacitor that  can  optionally  be  placed  either 


428  Antennas  and  Antenna  Matching 


THE  RADIO 


Figure  2 
ALTERNATIVE 
METHODS  OF 
FEEDING  A 
HALF-WAVE  DIPOLE 


CENTER-FED  TYPES 


THE  RADIO 


Multi - wire  Doublets  429 


in  series  with  the  antenna  coil  or  in  parallel 
with  it.  A series  tuning  capacitor  can  be 
placed  in  series  with  one  feeder  leg  without 
unbalancing  the  system. 

The  tuned-doublet  antenna  is  shown  in  fig- 
ure 2D.  The  antenna  is  a current-fed  system 
when  the  radiating  wire  is  a half  wave  long 
electrically,  or  when  the  system  is  operated 
on  its  odd  harmonics,  but  becomes  a voltage- 
fed  radiator  when  operated  on  its  even  har- 
monics. 

The  antenna  has  a different  radiation  pat- 
tern when  operated  on  its  harmonics,  as  would 
be  expected.  The  arrangement  used  on  the 
second  harmonic  is  better  known  as  the  Prank- 
lin  colinear  array  and  is  described  in  Chapter 
Twenty-three.  The  pattern  is  similar  to  a J^-wave 
dipole  except  that  it  is  sharper  in  the  broad- 
side direction.  On  higher  harmonics  of  oper- 
ation there  will  be  multiple  lobes  of  radiation 
from  the  system. 

Figures  2E  and  2F  show  alternative  arrange- 
ments for  using  an  untuned  transmission  line 
between  the  transmitter  and  the  tuned-doublet 
radiator.  In  figure  2E  a half-wave  shorted  line 
is  used  to  resonate  the  radiating  system, 
while  in  figure  2F  a quarter-wave  open  line  is 
utilized.  The  adjustment  of  quarter-wave  and 
half-wave  stubs  is  discussed  in  Section  19-8. 

Doublets  with  The  average  value  of  feed  im- 
Quorter-Wave  pedance  for  a center-fed  half- 
Transformers  wave  doublet  is  75  ohms.  The 
actual  value  varies  with  height 
and  is  shown  in  Chapter  Twenty-one.  Other 
methods  of  matching  this  rather  low  value  of 
impedance  to  a medium-impedance  transmis- 
sion line  are  shown  in  (G),  (H),  and  (I)  of  fig- 
ure 2.  Each  of  these  three  systems  uses  a 
quartet-wave  transformer  to  accomplish  the 
impedance  transformation.  The  only  difference 
between  the  three  systems  lies  in  the  type  of 
transmission  line  used  in  the  quarter-wave 
transformer.  (G)  shows  the  Johnson  Q system 
whereby  a line  made  up  of  !^-inch  dural  tubing 
is  used  for  the  low-impedance  linear  trans- 
former. A line  made  up  in  this  manner  is  fre- 
quently called  a set  of  Q bars.  Illustration 
(H)  shows  the  use  of  a four-wire  line  as  the 
linear  transformer,  and  (I)  shows  the  use  of  a 
piece  of  150-ohm  Twin-Lead  electrically  54- 
wave  in  length  as  the  transformer  between  the 
center  of  the  dipole  and  a piece  of  300-ohra 
Twin-Lead.  In  any  case  the  impedance  of  the 
quarter-wave  transformer  will  be  of  the  order 
of  150  to  200  ohms.  The  use  of  sections  of 
transmission  line  as  linear  transformers  is 
discussed  in  detail  in  Section  22-8. 

Multi-Wire  An  alternative  method  for  incteas- 
Doublets  ing  the  feed-point  impedance  of  a 
dipole  so  that  a medium-imped- 


ance transmission  line  may  be  used  is  shown 
in  figures  2J  and  2K.  This  system  utilizes 
more  than  one  wire  in  parallel  for  the  radiating 
element,  but  only  one  of  the  wires  is  broken 
for  attachment  of  the  feeder.  The  most  com- 
mon arrangement  uses  two  wires  in  the  flat 
top  of  the  antenna  so  that  an  impedance  multi- 
plication of  four  is  obtained. 

The  antenna  shown  in  figure  2J  is  the  so- 
called  Twin-Lead  folded  dipole  which  is  a 
commonly  used  antenna  system  on  the  medium- 
frequency  amateur  bands.  In  this  arrangement 
both  the  antenna  and  the  transmission  line  to 
the  transmitter  are  constructed  of  300-ohm 
Twin-Lead.  The  flat  top  of  the  antenna  is 
made  slightly  less  than  the  conventional 
length  (462/Fmc.  instead  of  468/F)jc.  for  a 
single-wire  flat  top)  and  the  two  ends  of  the 
Twin-Lead  are  joined  together  at  each  end. 
The  center  of  one  of  the  conductors  of  the 
Twin-Lead  flat  top  is  broken  and  the  two  ends 
of  the  Twin-Lead  feeder  are  spliced  into  the 
flat  top  leads.  As  a protection  against  mois- 
ture pieces  of  flat  polyethylene  taken  from 
another  piece  of  300-ohm  Twin-Lead  may  be 
molded  over  the  joint  between  conductors  with 
the  aid  of  an  electric  iron  or  soldering  iron. 

Better  bandwidth  characteristics  can  be  ob- 
tained with  a folded  dipole  made  of  ribbon  line 
if  the  two  conductors  of  the  ribbon  line  are 
shorted  a distance  of  0.82  (the  velocity  factor 
of  ribbon  line)  of  a free-space  quarter  wave- 
length from  the  center  or  feed  point.  This  pro- 
cedure is  illustrated  in  figure  3A.  An  alter- 
native arrangement  for  a Twin-Lead  folded 
dipole  is  illustrated  in  figure  3B.  This  type  of 
half-wave  antenna  system  is  convenient  for 
use  on  the  3.5-Mc.  band  when  the  116  to  132 
foot  distance  required  for  a full  half-wave  is 
not  quite  available  in  a straight  line,  since  the 
single-wire  end  pieces  may  be  bent  away  or 
downward  from  the  direction  of  the  main  sec- 
tion of  the  antenna. 

Figure  2K  shows  the  basic  type  of  2-wite 
doublet  or  folded  dipole  wherein  the  radiating 
section  of  the  system  is  made  up  of  standard 
antenna  wire  spaced  by  means  of  feeder 
spreaders.  The  feeder  again  is  made  of  300- 
ohm  Twin-Lead  since  the  feed-point  imped- 
ance is  approximately  300  ohms,  the  same  as 
that  of  the  Twin-Lead  folded  dipole. 

The  folded-dipole  type  of  antenna  has  the 
broadest  response  characteristic  (greatest 
bandwidth)  of  any  of  the  conventional  half- 
wave antenna  systems  constructed  of  small 
wires  or  conductors.  Hence  such  an  antenna 
may  be  operated  over  the  greatest  frequency 
range  without  serious  standing  waves  of  any 
common  half-wave  antenna  type. 

The  increased  bandwidth  of  the  multi-wire 
doublet  type  of  radiator,  and  the  fact  that  the 
feed-point  resistance  is  increased  several 


4 30  Antennas  and  Antenna  Matching 


THE  RADIO 


Figure  3 

FOLDED  DIPOLE  WITH  SHORTING 
STRAPS 

The  impedance  match  and  bandwidth  char- 
acteristics of  a folded  dipole  maybe  improved 
by  shorting  the  two  wires  of  the  ribbon  a dis- 
tance out  from  the  center  equal  to  the  veloci- 
ty factor  of  the  ribbon  times  the  half-length 
of  the  dipole  as  shown  at  (A),  An  alternative 
arrangement  with  bent  down  ends  for  space 
conservation  Is  Illustrated  at  (B), 


times  over  the  radiation  tesistance  of  the  ele- 
ment, have  both  contributed  to  the  frequent 
use  of  the  multi-wire  radiator  as  the  driven 
element  in  a parasitic  antenna  array. 

Delta-Matched  These  two  types  of  radiat- 

Doublet  and  ing  elements  are  shown  in 

Standard  Doublet  figure  2L  and  figure  2M.  The 
delta-matched  doublet  is 
described  in  detail  in  Section  22-8  of  this 
chapter.  The  standard  doublet,  shown  in  fig- 
ure 2M,  is  fed  in  the  center  by  means  of  75- 
ohm  Twin-Lead,  either  the  transmitting  or  the 
receiving  type,  or  it  may  be  fed  by  means  of 
twisted-pair  feeder  or  by  means  of  parallel- 
wire  lamp-cord.  Any  of  these  types  of  feed 
line  will  give  an  approximate  match  to  the 
center  impedance  of  the  dipole,  but  the  75- 
ohm  Twin-Lead  is  far  to  be  preferred  over  the 
other  types  of  low-impedance  feeder  due  to 
the  much  lower  losses  of  the  polyethylene- 
dielectric  transmission  line. 

The  coarial-cable-fed  doublet  shown  in  fig- 
ure 2N  is  a variation  on  the  system  shown  in 
figure  2M.  Either  52-ohm  coaxial  cable  or  75- 
ohm  coaxial  cable  may  be  used  to  feed  the 
center  of  the  dipole,  although  the  75-ohm  type 


Figure  4 

HALF-WAVE  VERTICAL  ANTENNA  SHOW- 
ING ALTERNATIVE  METHODS  OF  FEED 


will  give  a somewhat  b{/ter  impedance  match 
at  normal  antenna  heights.  Due  to  the  asym- 
metry of  the  coaxial  feed  system  difficulty 
may  be  encountered  with  waves  traveling  on 
the  outside  of  the  coaxial  cable.  For  this  rea- 
son the  use  of  Twin-Lead  is  normally  to  be 
preferred  over  the  use  of  coaxial  cable  for 
feeding  the  center  of  a half-wave  dipole. 

Off-Center  The  system  shown  in  figure 
Fed  Doublet  2(0)  is  sometimes  used  to 
feed  a half-wave  dipole,  espe- 
cially when  it  is  desired  to  use  the  same  an- 
tenna on  a number  of  harmonically-related  fre- 
quencies. The  feeder  wire  (no.  14  enamelled 
wire  should  be  used)  is  tapped  a distance  of 
14  per  cent  of  the  total  length  of  the  antenna 
either  side  of  center.  The  feeder  wire,  operat- 
ing against  ground  for  the  return  current,  has 
an  impedance  of  approximately  600  ohms.  The 
system  works  well  over  highly  conducting 
ground,  but  will  introduce  rather  high  losses 
when  the  antenna  is  located  above  rocky  or 
poorly  conducting  soil.  The  off-center  fed  an- 
tenna has  a further  disadvantage  that  it  is 
highly  responsive  to  harmonics  fed  to  it  from 
the  transmitter. 

The  effectiveness  of  the  antenna  system  in 
radiating  harmonics  is  of  course  an  advantage 
when  operation  of  the  antenna  on  a number  of 
frequency  bands  is  desired.  But  it  is  neces- 
sary to  use  a harmonic  filter  to  insure  that 
only  the  desired  frequency  is  fed  from  the 
transmitter  to  the  antenna. 


22-3  The  Half-Wave 

Vertical  Antenna 


The  half-wave  vertical  antenna  with  its  bot- 
tom end  from  0,1  to  0.2  wavelength  above 


handbook 


Vertical  Antennas  431 


Figure  5 

THE  LOW-FREQUENCY  GROUND  PLANE 
ANTENNA 

The  radiols  of  tho  ground  plane  antenna 
should  lie  in  a horizontal  plane,  although 
slight  departures  from  this  caused  by  nearby 
objects  is  allowable.  The  whip  may  be 
mounted  on  a short  post,  or  on  the  roof  of  a 
building.  The  wire  radials  may  slope  down~ 
wards  towards  their  tips,  acting  as  guy 
wires  for  the  installation. 


ground  is  an  effective  transmitting  antenna  for 
low-angle  radiation,  where  ground  conditions 
in  the  vicinity  of  the  antenna  are  good.  Such 
an  antenna  is  not  good  for  short-range  sky- 
wave  communication,  such  as  is  the  normal 
usage  of  the  3.5-Mc.  amateur  band,  but  is  ex- 
cellent for  short-range  ground-wave  communi- 
cation such  as  on  the  standard  broadcast  band 
and  on  the  amateur  1.8-Mc.  band.  The  vertical 
antenna  normally  will  cause  greater  BCI  than 
an  equivalent  horizontal  antenna,  due  to  the 
much  greater  ground-wave  field  intensity.  Al- 
so, the  vertical  antenna  is  poor  for  receiving 
under  conditions  where  man-made  interference 
is  severe,  since  such  interference  is  predomi- 
nantly vertically  polarized. 

Three  ways  of  feeding  a half-wave  vertical 
antenna  from  an  untuned  transmission  line  are 
illustrated  in  figure  4.  The  J-fed  system  shown 
in  figure  4A  is  obviously  not  practicable  ex- 
cept on  the  higher  frequencies  where  the  ex- 
tra length  for  the  stub  may  easily  be  obtained. 
However,  in  the  normal  case  the  ground-plane 
vertical  antenna  is  to  be  recommended  over 
the  J-fed  system  for  high  frequency  work. 

22-4  The  Ground  Plane  Antenna 

An  effective  low  angle  radiator  for  any  ama- 


Figure 6 

80  METER  LOADED  GROUND  PLANE 
ANTENNA 

Number  of  turns  in  loading  coil  to  be  adjusted 
until  antenna  system  resonates  at  desired 
frequency  in  80  meter  band. 


teur  band  is  the  ground-plane  antenna,  shown 
in  figure  5.  So  called  because  of  the  radial 
ground  wires,  the  ground-plane  antenna  is  not 
affected  by  soil  conditions  in  its  vicinity  due 
to  the  creation  of  an  artificial  ground  system 
by  the  radial  wires.  The  base  impedance  of 
the  ground  plane  is  of  the  order  of  30  to  35 
ohms,  and  it  may  be  fed  with  52-ohm  coaxial 
line  with  oidy  a slight  impedance  mis-match. 
For  a more  exact  match,  the  ground-plane  an- 
tenna may  be  fed  with  a 72-ohm  coaxial  line 
and  a quarter-wave  matching  section  made  of 
52-ohm  coaxial  line. 

The  angle  of  radiation  of  the  ground-plane 
antenna  is  quite  low,  and  the  antenna  will  be 
found  less  effective  for  contacts  under  1000 
miles  or  so  on  the  80  and  40  meter  bands  than 
a high  angle  radiator,  such  as  a dipole.  How- 
ever, for  DX  contacts  of  1000  miles  or  more, 
the  ground-plane  antenna  will  prove  to  be 
highly  effective. 

The  80-Meter  A vertical  antenna  of  66  feet 
Loaded  in  height  presents  quite  a prob- 

Ground-Plane  lem  on  a small  lot,  as  the  sup- 
porting guy  wires  will  tend  to 
take  up  quite  a large  portion  of  the  lot.  Under 
such  conditions,  it  is  possible  to  shorten  the 
length  of  the  vertical  radiator  of  the  ground- 
plane  by  the  inclusion  of  a loading  coil  in  the 
vertical  whip  section.  The  ground-plane  an- 
tenna may  be  artificially  loaded  in  this  man- 
ner so  that  a 25-foot  vertical  whip  may  be 


43  2 Antennas  and  Antenna  Matching 


the  radio 


used  for  the  radiator.  Such  an  antenna  is 
shown  in  figure  6.  The  loaded  ground-plane 
tends  to  have  a rather  high  operating  Q and 
operates  only  over  a narrow  band  of  frequen- 
cies. An  operating  range  of  about  100  kilo- 
cycles with  a low  SWR  is  possible  on  80  me- 
ters. Operation  over  a larger  frequency  range 
is  possible  if  a. higher  standing  wave  ratio  is 
tolerated  on  the  transmission  line.  The  radia- 
tion resistance  of  a loaded  80-meter  ground- 
plane  is  about  15  ohms.  A quartet  wavelength 
(45  feet)  of  52-ohm  coaxial  line  will  act  as  an 
efficient  feed  line,  presenting  a load  of  ap- 
proximately 180  ohms  to  the  transmitter. 


22-5  The  Marconi 

Antenna 


A grounded  quarter-wave  Marconi  antenna, 
widely  used  on  frequencies  below  3 Me.,  is 
sometimes  used  on  the  3.5-Mc.  band,  and  is 
also  used  in  v-h-f  mobile  services  where  a 
compact  antenna  is  required.  The  Marconi  type 
antenna  allows  the  use  of  half  the  length  of 
wire  that  would  be  required  for  a half-wave 
Hertz  radiator.  The  ground  acts  as  a mirror, 
in  effect,  and  takes  the  place  of  the  additional 
quarter-wave  of  wire  that  would  be  required 
to  reach  resonance  if  the  end  of  the  wire  were 
not  returned  to  ground. 

The  fundamental  practical  form  of  the  Mar- 
coni antenna  system  is  shown  in  figure  7. 
Other  Marconi  antennas  differ  from  this  type 
primarily  in  regard  to  the  method  of  feeding 
the  energy  to  the  radiator.  The  feed  method 
shown  in  figure  7B  can  often  be  used  to  advan- 
tage, particularly  in  mobile  work. 

Variations  on  the  basic  Marconi  antenna 
are  shown  in  the  illustrations  of  figure  8.  Fig- 
ures 8B  and  8C  show  the  "L”-type  and  "T”- 
type  Marconi  antennas.  These  arrangements 
have  been  more  or  less  superseded  by  the  top- 
loaded  forms  of  the  Marconi  antenna  shown  in 
figures  8D,  8E,  and  8F.  In  each  of  these  lat- 
ter three  figures  an  antenna  somewhat  less 
than  one  quartet  wave  in  length  has  been 
loaded  to  increase  its  effective  length  by  the 
insertion  of  a loading  coil  at  or  near  the  top 
of  the  radiator.  The  arrangement  shown  at  fig- 
ure 8D  gives  the  least  loading  but  is  the  most 
practical  mechanically.  The  system  shown  at 
figure  8E  gives  an  intermediate  amount  of 
loading,  while  that  shown  at  figure  8F,  utiliz- 
ing a "hat”  just  above  the  loading  coil,  gives 
the  greatest  amount  of  loading.  The  object  of 
all  the  top-loading  methods  shown  is  to  pro- 
duce an  increase  in  the  effective  length  of 
the  radiator,  and  thus  to  raise  the  point  of 
maximum  current  in  the  radiator  as  far  as  pos- 


© 


COAX.  FROM  TRANS. 


-j: 


Figure  7 

FEEDING  A QUARTER-WAVE  MARCONI 
ANTENNA 

When  on  open-wire  line  is  to  be  used^  it  may 
be  link  coupled  to  a series-resonant  circuit 
between  the  bottom  end  of  the  Marconi  and 
ground,  as  at  (A).  Alternatively,  a reason- 
ably good  impedance  match  may  be  obtained 
between  52-obm  coaxial  line  and  the  bottom 
of  a resonant  quarter-wave  antenna,  as  illus- 
trated at  (B)  above. 


sible  above  ground.  Raising  the  maximum-cur- 
rent point  in  the  radiator  above  ground  has 
two  desirable  results:  The  percentage  of  low- 
angle  radiation  is  increased  and  the  amount  of 
ground  current  at  the  base  of  the  radiator  is 
reduced,  thus  reducing  the  ground  losses. 

To  estimate  whether  a loading  coil  will 
probably  be  required,  it  is  necessary  only  to 
note  if  the  length  of  the  antenna  wire  and 
ground  lead  is  over  a quarter  wavelength;  if 
so,  no  loading  coil  is  needed,  provided  the 
series  tuning  capacitor  has  a high  maximum 
capacitance. 

Amateurs  primarily  interested  in  the  higher 
frequency  bands,  but  who  like  to  work  80  me- 
ters occasionally,  can  usually  manage  to  reso- 
nate one  of  their  antennas  as  a Marconi  by 
working  the  whole  system,  feeders  and  all, 
against  a water  pipe  ground,  and  resorting  to 
a loading  coil  if  necessary.  A high-frequency- 
rotary,  zepp,  doublet,  or  single-wire-fed  an- 
tenna will  make  quite  a good  80-meter  Marconi 
if  high  and  in  the  clear,  with  a rather  long 
feed  line  to  act  as  a radiator  on  80  meters. 
Where  two-wire  feeders  are  used,  the  feeders 
should  be  tied  together  for  Marconi  operation. 

Importance  of  With  a quarter-wave  anten- 

Ground  Connection  na  and  a ground,  the  an- 
tenna current  generally  is 
measured  with  a meter  placed  in  the  antenna 
circuit  close  to  the  ground  connection.  If  this 


HANDBOOK 


M arconi  Antenn  a 433 


Figure  8 
LOADING  THE 
MARCONI  ANTENNA 

The  various  loading  systems 
are  discussed  in  the  accom- 
panying text. 


r 

LOADING 

COILS  *HAT» 

;>  T 

LESS 

1-  J. 

THAN 

■5- 

© ® © © 


current  flows  through  a resistor,  or  if  the 
ground  itself  presents  some  resistance,  there 
will  be  a power  loss  in  the  form  of  heat.  Im- 
proving the  ground  connection,  therefore,  pro- 
vides a definite  means  of  reducing  this  power 
loss,  and  thus  increasing  the  radiated  power. 

The  best  possible  ground  consists  of  as 
many  wires  as  possible,  each  at  least  a quar- 
ter wave  long,  buried  just  below  the  surface 
of  the  earth,  and  extending  out  from  a common 
point  in  the  form  of  radials.  Copper  wire  of 
any  size  larger  than  no.  16  is  satisfactory, 
though  the  larger  sizes  will  take  longer  to  dis- 
integrate. In  fact,  the  radials  need  not  even 
be  buried;  they  may  be  supported  just  above 
the  earth,  and  insulated  from  it.  This  arrange- 
ment is  called  a counterpoise,  and  operates 
by  virtue  of  its  high  capacitance  to  ground. 

If  the  antenna  is  physically  shorter  than  a 
quarter  wavelength,  the  antenna  current  is 
higher,  due  to  lower  radiation  resistance.  Con- 
sequently, the  power  lost  in  resistive  soil  is 
greater.  The  importance  of  a good  ground  with 
short,  inductive-loaded  Marconi  radiators  is, 
therefore,  quite  obvious.  With  a good  ground 
system,  even  very  short  (one-eighth  wave- 
length) antennas  can  be  expected  to  give  a 
high  percentage  of  the  efficiency  of  a quarter- 
wave  antenna  used  with  the  same  ground  sys- 
tem. This  is  especially  true  when  the  short 
radiator  is  top  loaded  with  a high  Q (low  loss) 
coil. 

Water-Pipe  Water  pipe,  because  of  its  com- 
Grounds  paratively  large  surface  and  cross 
section,  has  a relatively  low  r-f 
resistance.  If  it  is  possible  to  attach  to  a 
junction  of  several  water  pipes  (where  they 
branch  in  several  directions  and  run  for  some 
distance  under  ground),  a satisfactory  ground 
connection  will  be  obtained.  If  one  of  the 
pipes  attaches  to  a lawn  or  garden  sprinkler 
system  in  the  immediate  vicinity  of  the  anten- 
na, the  effectiveness  of  the  system  will  ap- 
proach that  of  buried  copper  radials. 

The  main  objection  to  water-pipe  grounds 


is  the  possibility  of  high  resistance  joints  in 
the  pipe,  due  to  the  "dope”  put  on  the  cou- 
pling threads.  By  attaching  the  ground  wire 
to  a junction  with  three  or  more  legs,  the  pos- 
sibility of  requiring  the  main  portion  of  the 
t-f  current  to  flow  through  a high  resistance 
connection  is  greatly  reduced. 

The  presence  of  water  in  the  pipe  adds 
nothing  to  the  conductivity;  therefore  it  does 
not  relieve  the  problem  of  high  resistance 
joints.  Bonding  the  joints  is  the  best  insur- 
ance, but  this  is,  of  course,  impracticable 
where  the  pipe  is  buried.  Bonding  together 
with  copper  wire  the  various  water  faucets 
above  the  surface  of  the  ground  will  improve 
the  effectiveness  of  a water-pipe  ground  sys- 
tem hampered  by  high-resistance  pipe  cou- 
plings. 

Marconi  A Marconi  antenna  is  an  odd 

Dimensions  number  of  electrical  quarter 
waves  long  (usually  only  one 
quarter  wave  in  length),  and  is  always  reso- 
nated to  the  operating  frequency.  The  correct 
loading  of  the  final  amplifier  is  accomplished 
by  varying  the  coupling,  rather  than  by  detun- 
ing the  antenna  from  resonance. 

Physically,  a quarter-wave  Marconi  may  be 
made  anywhere  from  one-eighth  to  three-eighths 
wavelength  overall,  meaning  the  total  length  of 
the  antenna  wire  and  ground  lead  from  the  end 
of  the  antenna  to  the  point  where  the  ground 
lead  attaches  to  the  junction  of  the  radials  or 
counterpoise  wires,  or  where  the  water  pipe 
enters  the  ground.  The  longer  the  antenna 
is  made  physically,  the  lower  will  be  the  cur- 
rent flowing  in  the  ground  connection,  and  the 
greater  will  be  the  overall  radiation  efficiency. 
However,  when  the  antenna  length  exceeds 
three-eighths  wavelength,  the  antenna  be- 
comes difficult  to  resonate  by  means  of  a 
series  capacitor,  and  it  begins  to  take  shape 
as  an  end-fed  Hertz,  requiring  a method  of 
feed  such  as  a pi  network. 

A radiator  physically  much  shorter  than  a 


4 34  Antennas  and  Antenna  Matching 


THE  RADIO 


H X/4  AT  LOWEST  FREQUENCY- 


Flgure  9 

THREE  EFFECTIVE  SPACE  CONSERVING 
ANTENNAS 

The  arrangements  shown  at  (A)  and  (B)  are 
satisfactory  where  resonant  feed  line  can  be 
used.  However,  non-resonant  75-ohm  feed 
line  may  be  used  in  the  arrangement  at  (A) 
when  the  dimensions  in  wavelengths  are  as 
shown.  In  the  arrangement  shown  at  (B)  low 
standing  waves  will  be  obtained  on  the  feed 
line  when  the  overall  length  of  the  antenna 
is  a half  wave.  The  arrangement  shown  at 
(C)  may  be  tuned  for  any  reasonable  length 
of  flat  top  to  give  a minimum  of  standing 
waves  on  the  transmission  line. 


quarter  wavelength  can  be  lengthened  elec- 
trically by  means  of  a series  loading  coil,  and 
used  as  a quarter-wave  Marconi.  However,  if 
the  wire  is  made  shorter  than  approximately 
one-eighth  wavelength,  the  radiation  resist- 
ance will  be  quite  low.  This  is  a special  prob- 
lem in  mobile  work  below  about  20-Mc. 


22-6  Space-Conserving 

Antennas 


In  many  cases  it  is  desired  to  undertake  a 
considerable  amount  of  operation  on  the  80- 
meter  or  40-meter  band,  but  sufficient  space 
is  simply  not  available  for  the  installation  of 
a half-wave  radiator  for  the  desired  frequency 
of  operation.  This  is  a common  experience  of 
apartment  dwellers.  The  shortened  Marconi 
antenna  operated  against  a good  ground  can 
be  used  under  certain  conditions,  but  the  short- 
ened Marconi  is  notorious  for  the  production 
of  broadcast  interference,  and  a good  ground 
connection  is  usually  completely  unobtainable 
in  an  apartment  house. 


Figure  10 

TWIN-LEAD  MARCONI  ANTENNA  FOR  THE 
80  AND  160  METER  BANDS 


Essentially,  the  problem  in  producing  an 
antenna  for  lower  frequency  operation  in  re- 
stricted space  is  to  erect  a short  radiator 
which  is  balanced  with  respect  to  ground  and 
which  is  therefore  independent  of  ground  for 
its  operation.  Several  antenna  types  meeting 
this  set  of  conditions  are  shown  in  figure  9. 
Figure  9A  shows  a conventional  center-fed 
doublet  with  bent-down  ends.  This  type  of  an- 
tenna can  be  fed  with  75-ohm  Twin-Lead  in  the 
center,  or  it  may  be  fed  with  a resonant  line 
for  operation  on  several  bands.  The  overall 
length  of  the  radiating  wire  will  be  a few  per 
cent  greater  than  the  normal  length  for  such 
an  antenna  since  the  wire  is  bent  at  a posi- 
tion intermediate  between  a current  loop  and 
a voltage  loop.  The  actual  length  will  have  to 
be  determined  by  the  cut-and-try  process  be- 
cause of  the  increased  effect  of  interfering  ob- 
jects on  the  effective  electrical  length  of  an 
antenna  of  this  type. 

Figure  9B  shows  a method  for  using  a two- 
wire  doublet  on  one  half  of  its  normal  operat- 
ing frequency.  It  is  recommended  that  spaced 
open  conductor  be  used  both  for  the  radiating 
pwtion  of  the  folded  dipole  and  for  the  feed 
line.  The  reason  for  this  recommendation  lies 
in  the  fact  that  the  two  wires  of  the  flat  top 
are  not  at  the  same  potential  throughout  their 
length  when  the  antenna  is  operated  on  one- 
half  frequency.  Twin-Lead  may  be  used  for 
the  feed  line  if  operation  on  the  frequency 
where  the  flat  top  is  one-half  wave  in  length 
is  most  common,  and  operation  on  one-half  fre- 
quency is  infrequent.  However,  if  the  antenna 
is  to  be  used  primarily  on  one-half  frequency 
as  shown,  it  should  be  fed  by  means  of  an 
open-wire  line.  If  it  is  desired  to  feed  the  an- 
tenna with  a non-resonant  line,  a quarter-wave 
stub  may  be  connected  to  the  antenna  at  the 
points  X,  X in  figure  9B.  The  stub  should  be 
tuned  and  the  transmission  line  connected  to 
it  in  the  normal  manner. 

The  antenna  system  shown  in  figure  9C  may 
be  used  when  not  quite  enough  length  is  avail- 
able for  a full  half-wave  radiator.  The  dimen- 


HANDBOOK 


Space  Conserving  Antennas  435 


FOR  detail  see  FIG.  A 


PHENOLIC  BLOCK  2' X 1.5' X 
WRAP  CABLES  AND  BLOCK 
WITH  SCOTCH  ELECTRICALT^ 
SPACE  BLOCKS  6'  APART 
ALONG  BALUN 


FIGURE  A 

CUT  OFF  SHIELD  AND  OUTER 
JACKET  AS  SHOWN.  ALLOW 
DIELECTRIC  TO  EXTEND  PART 
WAY  TO  OTHER  CABLE.  COVER 
ALL  EXPOSED  SHIELD  AND 
DIELECTRIC  ON  BOTH  CABLES 
WITH  A CONTINUOUS  WRAP- 
PING OF  SCOTCH  ELECTRICAL 
TAPE  TO  EXCLUDE  MOISTURE. 


FOR  DETAIL  SEE  FIGURE  B 
S2  OHM  RG-e/U,  ANY  LENGTH 


FIGURE  B 

REMOVE  OUTER  JACKET 
FROM  A SHORT  LENGTH  OF 
CABLE  AS  SHOWN  HERE. 
UNBRAIO  THE  SHIELD  OF 
COAXC,  CUT  OFF  THE  DI- 
ELECTRIC AND  INNER  CON- 
DUCTOR FLUSH  WITH  THE 
OUTER  JACKET.  DO  NOT  CUT 
^ THE  SHIELD.  WRAPSHIELD 
-1  OF  COAX  C AROUND  SHIELD 
OF  COAX  0.  SOLDER  THE 
CONNECTION,  BEING  VERY 
CAREFUL  NOT  TO  DAMAGE 
THE  DIELECTRIC  MATERIAL. 
HOLD  CABLE  D STRAIGHT 

WHILE  SOLDERING.  COVER 

THE  AREA  WITH  A CONTIN- 

UOUS WRAPPING  OF  SCOTCH 
ELECTRICAL  TAPE.  NO  CON- 

NECTION TO  INNER  CONDUC- 
TORS. 


FOR  DETAIL  SEE  FIG.  A 


PHENOLIC  BLOCK  2'XI.S'X0.5*- 

WRAP  CABLES  AND  BLOCK  , 

WITH  SCOTCH  ELECTRICALTAreJ 

k SPACE  BLOCKS  «' APART 
^ALONG  BALUN 


m 

V//// 


THE  TWO  WIRES  MAYBE 
I SPREAD  EITHER  HORIZ- 
ONTALLY OR  VERTICALLY. 


PP 


Ki.s'J 


FIGURE  A 

CUT  OFF  SHIELD  AND  OUTER 
JACKET  AS  SHOWN.  ALLOW 
DIELECTRIC  TO  EXTEND  PART 
WAY  TO  OTHER  CABLE.  COVER 
ALL  EXPOSED  SHIELD  AND 
DIELECTRIC  ON  BOTH  CABLES 
WITH  A CONTINUOUS  WRAP- 

PING OF  SCOTCH  ELECTRICAL 
TAPE  TO  EXCLUDE  MOISTURE. 


1 


FOR  DETAIL  SEE  FIGURE  B- 
S2  OHM  RG-S/U.  ANY  LENGTH- 


FIGURE  B 
REMOVE  OUTER  JACKET 
FROM  A SHORT  LENGTH  OF 
CABLE  AS  SHOWN  HERE. 
UNBRAID  THE  SHIELD  OF 
COAX  C,  CUTOFF  THE  DI- 
ELECTRIC AND  INNER  CON- 
DUCTOR FLUSH  WITH  THE 
OUTER  JACKET.  DO  NOT  CUT 
THE  SHIELD.  WRAPSHIELD 
lOF  COAXC  AROUND  SHIELD 
OF  COAX  D.  SOLDER  THE 
CONNECTION,  BEING  VERY 
CAREFUL  NOT  TO  DAMAGE 
THE  DIELECTRIC  MATERIAL. 
HOLD  CABLE  D STRAIGHT 
WHILE  SOLDERING.  COVER 
THE  AREAWITHACONTIN- 
UOUS  WRAPPING  OF  SCOTCH 
ELECTRICAL  TAPE.  NO  CON- 
NECTION TO  INNER  OONDUC- 
TORS. 


OIMBNSIONS  SHOWN  HERg  ARS  ROR  THE  40  METER  BAHO.  TH/S  AHT^ 
EHHA  MA  r BE  BUILT  ROR  OTHER  BANOS  BY  USIN9  OIMENSIONS  THAT 
ARE  MULTIPLES  OR  SUBMULT/PLES  OR  THE  O/MENS/ONS  SHOWN, 
BALUN  SPACING  IS  hS'ONALL  BANOS- 


OIMENSIONS  SHOWN  HERE  ARE  ROR  THE  00  METER  BAND.  THIS  ANT- 
ENNA MAY  BE  BUILT  ROR  OTHER  BANOS  BY  USING  DIMENSIONS  THAT 
ARE  MULTIPLES  OR  SUBMULTIPLES  OR  THE  DIMENSIONS  SHOWN. 
BALUN  SPACING  IS  I.S'ONALL  BANOS. 


Figure  1 1 

HALF-WAVE  ANTENNA  WITH  QUARTER- 
WAVE  UNBALANCED  TO  BALANCED 
TRANSFORMER  (BALUN)  FEED  SYSTEM 
FOR  40.METER  OPERATION 


Figure  12 

BROADBAND  ANTENNA  WITH  QUARTER- 
WAVE  UNBALANCED  TO  BALANCED 
TRANSFORMER  (BALUN)  FEED  SYSTEM 
FOR  80.METER  OPERATION 


sions  in  terms  of  frequency  are  given  on  the 
drawing.  An  antenna  of  this  type  is  93  feet 
long  for  operation  on  3600  kc.  and  86  feet  long 
for  operation  on  3900  kc.  This  type  of  antenna 
has  the  additional  advantage  that  it  may  be 
operated  on  the  7-Mc.  and  14-Mc.  bands,  when 
the  flat  top  has  been  cut  for  the  3-5-Mc.  band, 
simply  by  changing  the  position  of  the  short- 
ing bar  and  the  feeder  line  on  the  stub. 

A sacrifice  which  must  be  made  when  using 
a shortened  radiating  system,  as  for  example 
the  types  shown  in  figure  9,  is  in  the  band- 
width of  the  radiating  system.  The  frequency 
range  which  may  be  covered  by  a shortened 
antenna  system  is  approximately  in  proportion 
to  the  amount  of  shortening  which  has  been 
employed.  For  example,  the  antenna  system 
shown  in  figure  9C  may  be  operated  over  the 
range  from  3800  kc.  to  4000  kc.  without  ser- 
ious standing  waves  on  the  feed  line.  If  the 


antenna  had  been  made  full  length  it  would 
be  possible  to  cover  about  half  again  as  much 
frequency  range  for  the  same  amount  of  mis- 
match on  the  extremes  of  the  frequency  range. 

The  Twin-Lead  Much  of  the  power  loss  in 
Marconi  Antenna  the  Marconi  antenna  is  a re- 
sult of  low  radiation  resist- 
ance and  high  ground  resistance.  In  some 
cases,  the  ground  resistance  may  even  be 
be  higher  than  the  radiation  resistance,  caus- 
ing a loss  of  50  per  cent  or  more  of  the  trans- 
mitter power  output.  If  the  radiation  resistance 
of  the  Marconi  antenna  is  raised,  the  amount 
of  power  lost  in  the  ground  resistance  is  pro- 
portionately less.  If  a Marconi  antenna  is  made 
out  of  300  ohm  TV-type  ribbon  line,  as  shown 
in  figure  10,  the  radiation  resistance  of  the 
antenna  is  raised  from  a low  value  of  10  or  15 
ohms  to  a more  reasonable  value  of  40  to  60 


43  6 Antennas  and  Antenna  Matching 


the  radio 


Figure  1 3 

SWR  CURVE  OF  BO-METER  BROAD-BAND 
DIPOLE 


ohms.  The  ground  losses  are  now  reduced  by 
a factor  of  4.  In  addition,  the  antenna  may  be 
directly  fed  from  a 50-ohm  coaxial  line,  or  di- 
rectly from  the  unbalanced  output  of  a pi-  net- 
work transmitter. 

Since  a certain  amount  of  power  may  still 
be  lost  in  the  ground  connection,  it  is  still  of 
greatest  importance  that  a good,  low  resist- 
ance ground  be  used  with  this  antenna. 

The  Collins  Shown  in  figures  11  and  12 

Brood-bond  are  broad-band  dipoles  for 

Dipole  System  the  40  and  80  meter  amateur 
bands,  designed  by  Collins 
Radio  Co.  for  use  with  the  Collins  32V-3  and 
KW-1  transmitters.  These  fan-type  dipoles 
have  excellent  broad-band  response,  and  are 
designed  to  be  fed  with  a 52-ohm  unbalanced 
coaxial  line,  making  them  suitable  for  use  with 
many  of  the  other  modern  transmitters,  such 
as  the  Barket  and  Williamson  5100,  Johnson 
Ranger,  and  Viking.  The  antenna  system  con- 
sists of  a fan-type  dipole,  a balun  matching 
section,  and  a suitable  coaxial  feedline.  The 
Q of  the  half-wave  80  meter  doublet  is  low- 
ered by  decreasing  the  effective  length-to- 
diameter  ratio.  The  frequency  range  of  opera- 
tion of  the  doublet  is  increased  considerably 
by  this  change.  A typical  SWR  curve  for  the 
80  meter  doublet  is  shown  in  figure  13. 

The  balanced  doublet  is  matched  to  the  un- 
balanced coaxial  line  by  the  one-quarter  wave 
balun.  If  desired,  a shortened  balun  may  be 
used  (figure  14).  The  short  balun  is  capacity 
loaded  at  the  junction  between  the  balun  and 
the  broad-band  dipole. 


22-7  Multi-Band  Antennas 

The  availability  of  a multi-band  antenna  is 
a great  opetating  convenience  to  an  amateur 
station.  In  most  cases  it  will  be  found  best  to 
install  an  antenna  which  is  optimum  for  the 
band  which  is  used  for  the  majority  of  the 


-ANTENNA 


Figure  14 

SHORT  BALUN  FOR  40  AND  80  METERS 


available  operating  time,  and  then  to  have  an 
additional  multi-band  antenna  which  may  be 
pressed  into  service  for  operation  on  another 
band  when  propagation  conditions  on  the  most 
frequently  used  band  are  not  suitable.  Most 
amateurs  use,  or  plan  to  install,  at  least  one 
directive  array  for  one  of  the  higher-frequency 
bands,  but  find  that  an  additional  antenna 
which  may  be  used  on  tbe  3.5-Mc.  and  7.0-Mc. 
band,  or  even  up  through  the  28-Mc.  band  is 
almost  indispensable. 

The  choice  of  a multi-band  antenna  depends 
upon  a number  of  factors  such  as  the  amount 
of  space  available,  the  band  which  is  to  be 
used  for  the  majority  of  operating  with  the  an- 
tenna, the  radiation  efficiency  which  is  de- 
sired, and  the  type  of  antenna  tuning  network 
to  be  used  at  the  transmitter.  A number  of 
recommended  types  are  shown  in  the  next 
pages. 

The  ?i-Wove  Figure  15  shows  an  antenna 

Folded  Doublet  type  which  will  be  found  to 
be  very  effective  when  a 
moderate  amount  of  space  is  available,  when 
most  of  the  operating  will  be  done  on  one  band 
with  occasional  operation  on  the  second  har- 
monic. The  system  is  quite  satisfactory  for 
use  with  high-power  transmitters  since  a 600- 
ohm  non-resonant  line  is  used  from  the  anten- 
na to  the  transmitter  and  since  the  antenna 
system  is  balanced  with  respect  to  ground. 
With  operation  on  the  fundamental  frequency 
of  the  antenna  where  the  flat  top  is  54  wave 
long  the  switch  SW  is  left  open.  The  system 
affords  a very  close  match  between  tbe  600- 
ohm  line  and  the  feed  point  of  the  antenna. 
Kraus  has  reported  a standing-wave  ratio  of 
approximately  1.2  to  1 over  the  14-Mc.  band 
when  the  antenna  was  located  approximately 
one-half  wave  above  ground. 

For  operation  on  the  second  harmonic  the 
switch  SW  is  closed.  The  antenna  is  still  an 


HANDBOOK 


Multi-band  Antennas  437 


h = L ^ 


Figure  15 

THE  THREE-QUARTER  WAVE  FOLDED 
DOUBLET 

This  onfenno  arrangement  will  give  very 
satisfactory  operation  with  a SOO-ohm  feed 
line  for  operation  with  the  switch  open  on 
the  fundamental  frequency  and  with  the 
switch  closed  on  twice  frequency. 


effective  radiator  on  the  second  harmonic  but 
the  pattern  of  radiation  will  be  different  from 
that  on  the  fundamental,  and  the  standing-wave 
ratio  on  the  feed  line  will  be  greater.  The  flat 
top  of  the  antenna  must  be  made  of  open  wire 
rather  than  ribbon  or  tubular  line. 

For  greater  operating  convenience,  the  short- 
ing switch  may  be  replaced  with  a section  of 
transmission  line.  If  this  transmission  line  is 
made  one-quarter  wavelength  long  for  the  fun- 
damental frequency,  and  the  free  end  of  the 
line  is  shorted,  it  will  act  as  an  open  circuit 
across  the  center  insulator.  At  the  second  har- 
monic, the  transmission  line  is  one-half  wave- 
length long,  and  reflects  the  low  impedance 
of  the  shorted  end  across  the  center  insulator. 
Thus  the  switching  action  is  automatic  as  the 
frequency  of  operation  is  changed.  Such  an 
installation  is  shown  in  figure  16. 

The  End- Fed  The  end-fed  Hertz  antenna 
Hertz  shown  in  figure  17  is  not  as 

effective  a radiating  system  as 


Figure  17 

RECOMMENDED  LENGTHS  FOR  THE  END-.^ 
FED  HERTZ 


L = 33FT.  " " " 98  FT. 

L=  16.5  FT  • " " >19.6  FT. 


Figure  16 

AUTOMATIC  BANDSWITCHING  STUB  FOR 
THE  THREE-QUARTER  WAVE  FOLDED 
DOUBLET 

The  antenna  of  Figure  15  may  be  used  with 
a shorted  stub  line  in  place  of  the  switch 
normally  used  for  second  harmonic  operation. 


many  other  antenna  types,  but  it  is  particular- 
ly convenient  when  it  is  desired  to  install  an 
antenna  in  a hurry  for  a test,  or  for  field-day 
work.  The  flat  top  of  the  radiator  should  be 
as  high  and  in  the  clear  as  possible.  In  any 
event  at  least  three  quarters  of  the  total  wire 
length  should  be  in  the  clear.  Dimensions  for 
optimum  operation  on  various  amateur  bands 
are  given  in  addition  in  figure  17. 

The  End- Fed  The  end- fed  Zepp  has  long 
Zepp  been  a favorite  for  multi-band 

operation.  It  is  shown  in  fig- 
ure 18  along  with  recommended  dimensions 
for  operation  on  various  amateur  band  groups. 


END-FED  7EPP 


FIGURE  18 


4 38  Antennas  and  Antenna  Matching 


THE  RADIO 


Figure  19 

A TWO-BAND  MARCONI  ANTENNA  FOR 
160-80  METER  OPERATION 


Since  this  antenna  type  is  an  unbalanced  radi- 
ating system,  its  use  is  not  recommended  with 
high-power  transmitters  where  interference  to 
broadcast  listeners  Is  likely  to  be  encountered. 
The  r-f  voltages  encountered  at  the  end  of 
zepp  feeders  and  at  points  an  electrical  half 
wave  from  the  end  are  likely  to  be  quite  high. 
Hence  the  feeders  should  be  supported  an  ade- 
quate distance  from  surrounding  objects  and 
sufficiently  in  the  clear  so  that  a chance  en- 
counter between  a passerby  and  the  feeder  is 
unlikely. 

The  coupling  coil  at  the  transmitter  end  of 
the  feeder  system  should  be  link  coupled  to 
the  output  of  the  low-pass  TVI  filter  in  order 
to  reduce  harmonic  radiation. 

The  Two-Band  A three-eighths  wavelength 

Marconi  Antenna  Marconi  antenna  may  be 

operated  on  its  harmonic 
frequency,  providing  good  two  band  perform- 
ance from  a simple  wire.  Such  an  arrangement 
for  operation  on  160-80  meters,  and  80-40  me- 
ters is  shown  in  figure  19.  On  the  fundamental 
(lowest)  frequency,  the  antenna  acts  as  a 
three-eighths  wavelength  series-tuned  Marconi. 
On  the  second  harmonic,  the  antenna  is  a cut- 
tent-fed  three-quarter  wavelength  antenna  oper- 
ating against  ground.  For  proper  operation, 
the  antenna  should  be  resonated  on  its  second 
harmonic  by  means  of  a grid-dip  oscillator  to 
the  operating  frequency  most  used  on  this  par- 
ticular band.  The  Q of  the  antenna  is  relatively 
low,  and  the  antenna  will  perform  well  over  a 
frequency  range  of  several  hundred  kilocycles. 

The  overall  length  of  the  antenna  may  be 
varied  slightly  to  place  its  self-resonant  fre- 
quency in  the  desired  region.  Bends  or  turns 
in  the  antenna  tend  to  make  it  resonate  higher 
in  frequency,  and  it  may  be  necessary  to 
lengthen  it  a bit  to  resonate  it  at  the  chosen 
frequency.  For  fundamental  operation,  the 
series  condenser  is  inserted  in  the  circuit,  and 
the  antenna  may  be  resonated  to  any  point  in 
the  lower  frequency  band.  As  with  any  Marconi 


Li ^ 


CENTER-FED  ANTENNA 


Figure  20 

DIMENSIONS  FOR  CENTER-FED  MULTI- 
BAND ANTENNA 


type  antenna,  the  use  of  a good  ground  is  es- 
sential. This  antenna  works  well  with  trans- 
mitters employing  coaxial  antenna  feed,  since 
its  transmitting  impedance  on  both  bands  is  in 
the  neighborhood  of  40  to  60  ohms.  It  may  be 
attached  directly  to  the  output  terminal  of  such 
transmitters  as  the  Collins  32V  and  the  Viking 
11.  The  use  of  a low-pass  TVI  filter  is  of 
course  recommended. 

The  Center-Fed  For  multi-band  operation, 

Multi-Band  Antenna  the  center  fed  antenna  is 
without  doubt  the  best 
compromise.  It  is  a balanced  system  on  all 
bands,  it  requires  no  ground  return,  and  when 
properly  tuned  has  good  rejection  properties 
for  the  higher  harmonics  generated  in  the  trans- 
mitter. It  is  well  suited  for  use  with  the  various 
multi-band  150-watt  transmitters  that  are  cur- 
rently so  popular.  For  proper  operation  with 
these  transmitters,  an  antenna  tuning  unit 
must  be  used  with  the  center-fed  antenna.  In 
fact,  some  sort  of  tuning  unit  is  necessary  for 
any  type  of  efficient,  multi-band  antenna.  The 
use  of  such  questionable  antennas  as  the  "off- 
center  fed”  doublet  is  an  invitation  to  TVI 
troubles  and  improper  operation  of  the  trans- 
mitter. A properly  balanced  antenna  is  the 
best  solution  to  multi-band  operation.  When 
used  in  conjunction  with  an  antenna  tuning 
unit,  it  will  perform  with  top  efficiency  on  all 
of  the  major  amateur  bands. 

Several  types  of  center-fed  antenna  systems 
are  shown  in  figure  20.  If  the  feed  line  is  made 
up  in  the  conventional  manner  of  no.  12  or  no. 


HANDBOOK 


Multi-band  Antennas  439 


DIPOLE  TO  LIMIT  IMPEDANCE  EXCUR- 
SIONS ON  HARMONIC  FREQUENCIES 


14  wire  spaced  4 to  6 inches  the  antenna  sys- 
tem is  sometimes  called  a center-fed  zepp. 
With  this  type  of  feeder  the  impedance  at  the 
transmitter  end  of  the  feeder  varies  from  about 
70  ohms  to  approximately  5000  ohms,  the  same 
as  is  encountered  in  an  end-fed  zepp  antenna. 
This  great  impedance  ratio  requires  provision 
for  either  series  or  parallel  tuning  of  the  feed- 
ers at  the  transmitter,  and  involves  quite  high 
r-f  voltages  at  various  points  along  the  feed 
line. 

If  the  feed  line  between  the  transmitter  and 
the  antenna  is  made  to  have  a characteristic 
impedance  of  approximately  300  ohms  the  ex- 
cursions in  end-of-feeder  impedance  are  great- 
ly reduced.  In  fact  the  impedance  then  varies 
from  approximately  75  ohms  to  1200  ohms. 
With  this  much  lowered  impedance  variation 
it  is  usually  possible  to  use  series  tuning  on 
all  bands,  or  merely  to  couple  the  antenna  di- 
rectly to  the  output  tank  circuit  or  the  har- 
monic reduction  circuit  without  any  separate 
feeder  tuning  provision. 

There  are  several  practicable  types  of  trans- 
mission line  which  can  give  an  impedance  of 
approximately  300  ohms.  The  first  is,  obvious- 
ly, 300-ohm  Twin-Lead.  Twin-Lead  of  the  re- 
ceiving type  may  be  used  as  a resonant  feed 
line  in  this  case,  but  its  use  is  not  recom- 
mended with  power  levels  greater  than  perhaps 
150  watts,  and  it  should  not  be  used  when 
lowest  loss  in  the  transmission  line  is  desired. 

For  power  levels  up  to  250  watts  or  so,  the 
transmitting  type  tubular  300-ohm  line  may  be 
used,  or  the  open-wire  300-ohm  TV  line  may 
be  employed.  For  power  levels  higher  than 
this,  a 4-  wire  transmission  line.  Of  a line 
built  of  one-quarter  inch  tubing  should  be 
used. 


r- 


Figure  22 

FOLDED-TOP  DUAL-BAND  ANTENNA 


Even  when  a 300-ohm  transmission  line  is 
used,  the  end-of-feeder  impedance  may  reach 
a high  value,  particularly  on  the  second  har- 
monic of  the  antenna.  To  limit  the  impedance 
excursions,^  a two-wire  flat-top  may  be  em- 
ployed for  the  radiator,  as  shown  in  figure  21. 
The  use  of  such  a radiator  will  limit  the  im- 
pedance excursions  on  the  harmonic  frequen- 
cies of  the  antenna  and  make  the  operation  of 
the  antenna  matching  unit  much  less  critical. 
The  use  of  a two-wire  radiator  is  highly  recom- 
mended for  any  center-fed  multi-band  antenna. 

Folded  Flof-Top  As  has  been  mentioned 
Dual- Bond  Antenno  earlier,  there  is  an  increas- 
ing tendency  among  ama- 
teur operators  to  utilize  rotary  or  fixed  arrays 
for  the  14-Mc.  band  and  those  higher  in  fre- 
quency. In  order  to  afford  complete  coverage 
of  the  amateur  bands  it  is  then  desirable  to 
have  an  additional  system  which  will  operate 
with  equal  effectiveness  on  the  3.5-Mc.  and 
7-Mc.  bands,  but  this  low-frequency  antenna 
system  will  not  be  required  to  operate  on  any 
bands  higher  in  frequency  than  the  7-Mc.  band. 
The  antenna  system  shown  in  figure  22  has 
been  developed  to  fill  this  need. 

This  system  consists  essentially  of  an 
open-line  folded  dipole  for  the  7-Mc.  band  with 
a special  feed  system  which  allows  the  an- 
tenna to  be  fed  with  minimum  standing  waves 
on  the  feed  line  on  both  the  7-Mc.  and  3.5-Mc. 
bands.  The  feed-point  impedance  of  a folded 
dipole  on  its  fundamental  frequency  is  approxi- 
mately 300  ohms.  Hence  the  300-ohm  Twin- 
Lead  shown  in  figure  22  can  be  connected  di- 
rectly into  the  center  of  the  system  for  opera- 
tion only  on  the  7-Mc.  band  and  standing  waves 

on  the  feeder  will  be  very  small.  However,  it 
is  possible  to  insert  an  electrical  half-wave 


44  0 Antennas  and  Antenna  Matching 


THE  RADIO 


of  transmission  line  of  any  characteristic  im- 
pedance into  a feeder  system  such  as  this  and 
the  impedance  at  the  far  end  of  the  line  will 
be  exactly  the  same  value  of  impedance  which 
the  half-wave  line  sees  at  its  termination. 
Hence  this  has  been  done  in  the  antenna  sys- 
tem shown  in  figure  22;  an  electrical  half 
wave  of  line  has  been  inserted  between  the 
feed  point  of  the  antenna  and  the  300-ohm 
transmission  line  to  the  transmitter. 

The  characteristic  impedance  of  this  addi- 
tional half-wave  section  of  transmission  line 
has  been  made  about  715  ohms  (no.  20  wire 
spaced  6 inches),  but  since  it  is  an  electrical 
half  wave  long  at  7 Me.  and  operates  into  a 
load  of  300  ohms  at  the  antenna  the  300-ohm 
Twin-Lead  at  the  bottom  of  the  half-wave  sec- 
tion still  sees  an  impedance  of  300  ohms.  The 
additional  half-wave  section  of  transmission 
line  introduces  a negligible  amount  of  loss 
since  the  current  flowing  in  the  section  of  line 
is  the  same  which  would  flow  in  a 300-ohm 
line  at  each  end  of  the  half-wave  section,  and 
at  all  other  points  it  is  less  than  the  current 
which  would  flow  in  a 300-ohm  line  since  the 
effective  impedance  is  greater  than  300  ohms 
in  the  center  of  the  half-wave  section.  This 
means  that  the  loss  is  less  than  it  would  be  in 
an  equivalent  length  of  300-ohm  Twin-Lead 
since  this  type  of  manufactured  transmission 
line  is  made  up  of  conductors  which  ate  equiv- 
alent to  no.  20  wire. 

So  we  see  that  the  added  section  of  715-ohm 
line  has  substantially  no  effect  on  the  opera- 
tion of  the  antenna  system  on  the  7-Mc.  band. 
However,  when  the  flat  top  of  the  antenna  is 
operated  on  the  3.5-Mc.  band  the  feed-point 
impedance  of  the  flat  top  is  approximately 
3500  ohms.  Since  the  section  of  715-ohm  trans- 
mission line  is  an  electrical  quarter-wave  in 
length  on  the  3.5-Mc.  band,  this  section  of 
line  will  have  the  effect  of  transforming  the 
approximately  3500  ohms  feed-point  imped- 
ance of  the  antenna  down  to  an  impedance  of 
about  150  ohms  which  will  result  in  a 2:1 
standing- wave  ratio  on  the  300-ohm  Twin-Lead 
transmission  line  from  the  transmitter  to  the 
antenna  system. 

The  antenna  system  of  figure  22  operates 
with  very  low  standing  waves  over  the  entire 
7-Mc.  band,  and  it  will  operate  with  moderate 
standing  waves  from  3500  to  3800  kc.  in  the 
3.5-Mc.  band  and  with  sufficiently  low  stand- 
ing-wave ratio  so  that  it  is  quite  usable  over 
the  entire  3.5-Mc.  band. 

This  antenna  system,  as  well  as  all  other 
types  of  multi-band  antenna  systems,  must  be 
used  in  conjunction  with  some  type  of  har- 
monic-reducing antenna  tuning  network  even 
though  the  system  does  present  a convenient 
impedance  value  on  both  bands. 


Figure  23 

THE  MULTEE  TWO-BAND  ANTENNA 

Thiz  compact  antenna  can  be  uzed  with  ex- 
cellent resultz  on  760/80  and  80/40  meters. 
The  feedline  should  be  held  os  vertical  as 
possible,  since  it  radiates  when  the  antenna 
is  operated  on  its  fundamental  frequency. 


The  "Multee”  An  antenna  that  works  well 
Antenna  on  160  and  80  meters,  or  80 

and  40  meters  and  is  suffi- 
ciently compact  to  permit  erection  on  the  aver- 
age city  lot  is  the  W6BCX  Multee  antenna, 
illustrated  in  figure  23.  The  antenna  evolves 
from  a vertical  two  wire  radiator,  fed  on  one 
leg  only.  On  the  low  frequency  band  the  top 
portion  does  little  radiating,  so  it  is  folded 
down  to  form  a radiator  for  the  higher  frequen- 
cy band.  On  the  lower  frequency  band,  the  an- 
tenna acts  as  a top  loaded  vertical  radiator, 
while  on  the  higher  frequency  band,  the  flat- 
top  does  the  radiating  rather  than  the  vertical 
portion.  The  vertical  portion  acts  as  a quarter- 
wave  linear  transformer,  matching  the  6000 
ohm  antenna  impedance  to  the  50  ohm  imped- 
ance of  the  coaxial  transmission  line. 

The  earth  below  a vertical  radiator  must  be 
of  good  conductivity  not  only  to  provide  a low 
resistance  ground  connection,  but  also  to  pro- 
vide a good  reflecting  surface  for  the  waves 
radiated  downward  towards  the  ground.  For 
best  results,  a radial  system  should  be  in- 
stalled beneath  the  antenna.  For  160-80  me- 
ter operation,  six  radials  50  feet  in  length, 
made  of  no.  16  copper  wire  should  be  buried 
just  below  the  surface  of  the  ground.  While  an 
CM-dinary  water  pipe  ground  system  with  no 
radials  may  be  used,  a system  of  radials  will 
provide  a worthwhile  increase  in  signal 
strength.  For  80-40  meter  operation,  the  length 


HANDBOOK 


Low  Frequency  Discone  441 


20,I5.II.I0.6METERS 

D~  12'  L=  16' 

3*10"  Rs|8' 


DIMENSIONS 

15,11,10.6  METERS 
D=8'  L=I2' 

3*6"  Rs|2' 

HsiO'S" 


M,  10, 6, 2 METERS 
0*6'  L=9'6" 

S=4'  R=9'6' 
H=6'3" 


Figure  24 

DIMENSIONS  OF  LOW-FREQUENCY  DIS- 
CONE ANTENNA  FOR  LOW  FREQUENCY 
CUTOFF  AT  13.2  MC.,  20.1  MC,  AND 
26  MC. 

The  Discone  is  a verficaUy  polarized  radio- 
for,  producing  an  omnidirectional  pattern 
similar  to  a ground  plane.  Operation  on  sev- 
eral amateur  bands  with  low  SWR  on  the  co- 
axial feed  line  is  possible.  Additional  in- 
formation on  L-F  Discone  fay  W2RYI  in  July, 
1950  CQ  magazine. 


of  the  radials  may  be  reduced  to  25  feet.  As 
with  all  multi-band  antennas  that  employ  no 
lumped  tuned  circuits,  this  antenna  offers  no 
attenuation  to  harmonics  of  the  transmitter. 
When  operating  on  the  lower  frequency  band, 
it  would  be  wise  to  check  the  transmitter  for 
second  harmonic  emission,  since  this  antenna 
will  effectively  radiate  this  harmonic. 

The  Low-Frequency  The  discone  antenna  is 
Discone  widely  used  on  the  v-h-f 

bands,  but  until  recently 
it  has  not  been  put  to  any  great  use  on  the 
lower  frequency  bands.  Since  the  discone  is  a 
broad-band  device,  it  may  be  used  on  several 
harmonically  related  amateur  bands.  Size  is 
the  limiting  factor  in  the  use  of  a discone,  and 
the  20  meter  band  is  about  the  lowest  practi- 
cal frequency  for  a discone  of  reasonable  di- 
mensions. A discone  designed  for  20  meter 
operation  may  be  used  on  20,  15,  11,  10  and 


Figure  25 

SWR  CURVE  FOR  A 13.2  MC.  DISCONE 
ANTENNA.  SWR  IS  BELOW  1.5  TO  1 FROM 
13.0  MC.  TO  58  MC. 


6 meters  with  excellent  results.  It  affords  a 
good  match  to  a 50  ohm  coaxial  feed  system 
on  all  of  these  bands.  A practical  discone  an- 
tenna is  shown  in  figure  24,  with  a SWR  curve 
for  its  operation  over  the  frequency  range  of 
13-55  Me.  shown  in  figure  25.  The  discone 
antenna  radiates  a vertically  polarized  wave 
and  has  a very  low  angle  of  radiation.  For 
v-h-f  work  the  discone  is  constructed  of  sheet 
metal,  but  for  low  frequency  work  it  may  be 
made  of  copper  wire  and  aluminum  angle 
stock.  A suitable  mechanical  layout  for  a low 
frequency  discone  is  shown  in  figure  26. 
Smaller  versions  of  this  antenna  may  be  con- 
structed for  15,  11,  10  and  6 meters,  or  for  11, 
10,  6 and  2 meters  as  shown  in  the  chart  of 
figure  24. 

For  minimum  wind  resistance,  the  top  "hat” 
of  the  discone  is  constructed  from  three-quar- 
ter inch  aluminum  angle  stock,  the  rods  being 
bolted  to  an  aluminum  plate  at  the  center  of 
the  structure.  The  tips  of  the  tods  are  all  con- 
nected together  by  lengths  of  no.  12  enamelled 
copper  wire.  The  cone  elements  are  made  of 
no.  12  copper  wire  and  act  as  guy  wires  for 
the  discone  structure.  A very  rigid  arrange- 
ment may  be  made  from  this  design;  one  that 
will  give  no  trouble  in  high  winds.  A 4”  x 4" 
post  can  be  used  to  support  the  discone  struc- 
ture. 

The  discone  antenna  may  be  fed  by  a length 
of  50-ohm  coaxial  cable  directly  from  the  trans- 
mitter, with  a very  low  SWR  on  all  bands. 


The  Single-Wire-  The  old  favorite  single-wire- 
Fed  Antenna  fed  antenna  system  is  quite 
satisfactory  for  an  impromp- 
tu all  band  antenna  system.  It  is  widely  used 
for  portable  installations  and  "Field  Day” 
contests  where  a simple,  multi-band  antenna 
is  requited.  A single  wire  feeder  has  a char- 
acteristic impedance  of  some  500  ohms,  de- 


442  Antennas  and  Antenna  Matching 


THE  RADIO 


Figure  26 

MECHANICAL  CONSTRUCTION  OF  20- 
METER  DISCONE 


pending  upon  the  wire  size  and  the  point  of 
attachment  to  the  antenna.  The  earth  losses 
are  comparatively  low  over  ground  of  good  con- 
ductivity. Since  the  single  wire  feeder  radiates, 
it  is  necessary  to  bring  it  a^vay  from  the  an- 
tenna at  right  angles  to  the  antenna  wire 
for  at  least  one-half  the  length  of  the  an- 
tenna. 

The  correct  point  for  best  impedance  match 
on  the  fundamental  frequency  is  not  suitable 
for  harmonic  operation  of  the  antenna.  In  addi- 
tion, the  correct  length  of  the  antenna  for 
fundamental  operation  is  not  correct  for  har- 
monic operation.  Consequently,  a compromise 


CENTER 

I ANTENNA  WIRE 


FEEDER 


Figure  27 

SINGLE-WIRE-FED  ANTENNA  FOR  ALL- 
BAND OPERATION 

An  antenna  of  this  type  for  40-,  20.  and  JO. 
meter  operation  would  have  a radiator  67 
feet  long,  with  the  feeder  topped  7 7 feet  off 
center.  The  feeder  can  be  33,  66  or  99  feet 
tong.  The  some  type  of  antenna  for  80.,  40., 
20.  and  JO.meter  operation  would  have  a 
radiator  J34  feet  tong,  with  the  feeder  tapped 
22  feet  off  center.  The  feeder  can  be  either 
66  or  132  feet  long.  This  system  should  be 
used  only  with  those  coupling  methods  which 
provide  good  harmonic  suppression. 


must  be  made  in  antenna  length  and  point  of 
feeder  connection  to  enable  the  single-wire- 
fed  antenna  to  operate  on  more  than  one  band. 
Such  a compromise  introduces  additional  re- 
actance into  the  single  wire  feeder,  and  might 
cause  loading  difficulties  with  pi-network 
transmitters.  To  minimize  this  trouble,  the 
single  wire  feeder  should  be  made  a multiple 
of  33  feet  long. 

Two  typical  single-wire-fed  antenna  sys- 
tems are  shown  in  figure  27  with  dimensions 
for  multi-band  operation. 


22-8  Matching  Non-Resonant 
Lines  to  the  Antenna 

Present  practice  in  regard  to  the  use  of 
transmission  lines  for  feeding  antenna  systems 
on  the  amateur  bands  is  about  equally  divided 


HANDBOOK 


Matching  Systems  44  3 


NON-RESONANT 

LINE 


Figure  28 

THE  DELTA-MATCHED  DIPOLE 
ANTENNA 

7 he  dimensions  for  fhe  portions  of  the  an- 
tenna are  given  in  the  text. 


between  three  types  of  transmission  line:  (1) 
Ribbon  or  tubular  molded  300-ohm  line  is 
widely  used  up  to  moderate  power  levels  (the 
"transmitting”  type  is  useable  up  to  the  kilo- 
watt level).  (2)  Open-wire  400  to  600  ohm  line 
is  most  commonly  used  when  the  antenna  is 
some  distance  from  the  transmitter,  because  of 
the  low  attenuation  of  this  type  of  line.  (3)  Co- 
axial line  (usually  RG-8/U  with  a 52-ohm 
characteristic  impedance)  is  widely  used  in 
v-h-f  work  and  also  on  the  lower  frequencies 
where  the  feed  line  must  tun  underground  or 
through  the  walls  of  a building.  Coaxial  line 
also  is  of  assistance  in  TVI  reduction  since 
the  t-f  field  is  entirely  enclosed  within  the 
line.  Molded  75-ohm  line  is  sometimes  used 
to  feed  a doublet  antenna,  but  the  doublet  has 
been  largely  superseded  by  the  folded-dipole 
antenna  fed  by  300-ohm  ribbon  or  tubular  line 
when  an  antenna  for  a single  band  is  requited. 

Standing  Waves  As  was  discussed  earlier, 
standing  waves  on  the  anten- 
na transmission  line,  in  the  transmitting  case, 
ate  a result  of  reflection  from  the  point  where 
the  feed  line  joins  the  antenna  system.  The 
magnitude  of  the  standing  waves  is  deter- 
mined by  the  degree  of  mismatch  between  the 
characteristic  impedance  of  the  transmission 
line  and  the  input  impedance  of  the  antenna 
system.  When  the  feed-point  impedance  of  the 
antenna  is  resistive  and  of  the  same  value  as 
the  characteristic  impedance  of  the  feed  line, 
standing  waves  will  not  exist  on  the  feeder. 
It  may  be  well  to  repeat  at  this  time  that  there 
is  no  adjustment  which  can  be  made  at  the 


transmitter  end  of  the  feed  line  which  will 
change  the  magnitude  of  the  standing  waves 
on  the  anterma  transmission  line. 

Delta-Matched  The  delta  type  matched-im- 

Antenna  System  pedance  antenna  system  is 

shown  in  figure  28.  The  im- 
pedance of  the  transmission  line  is  trans- 
formed gradually  into  a higher  value  by  the 
fanned-out  Y portion  of  the  feeders,  and  the 
Y portion  is  tapped  on  the  antenna  at  points 
where  the  antenna  impedance  is  a compromise 
between  the  impedance  at  the  ends  of  the  Y 
and  the  impedance  of  the  unfanned  portion  of 
the  line. 

The  constants  of  the  system  ate  rather  criti- 
cal, and  the  antenna  must  resonate  at  the 
operating  frequency  in  order  to  minimize  stand- 
ing waves  on  the  line.  Some  slight  readjust- 
ment of  the  taps  on  the  antenna  is  desirable, 
if  appreciable  standing  waves  persist  in  ap- 
pearing on  the  line. 

The  constants  for  a doublet  ate  determined 
by  the  following  formulas: 

, 467.4 

i-teet  = 

F 

megacycles 


F 

megacycles 

147.6 

Efee.  - 

megacycles 

Where  L is  antenna  length;  D is  the  distance  in 
from  each  end  at  which  the  Y taps  on;  E is  the 
height  of  the  Y section. 

Since  these  constants  are  correct  only  for  a 
600-ohm  transmission  line,  the  spacing  S of 
the  line  must  be  approximately  75  times  the 
diameter  of  the  wire  used  in  the  transmission 
line.  For  no.  14  B 8t  S wire,  the  spacing  will 
be  slightly  less  than  5 inches.  This  system 
should  never  be  used  on  either  its  even  or  odd 
harmonics,  as  entirely  different  constants  are 
required  when  more  than  a single  half  wave- 
length appears  on  the  radiating  portion  of  the 
system. 

Multi-Wire  Doublets  When  a doublet  antenna 
or  the  driven  element  in 
an  array  consists  of  more  than  one  wire  or 
tubing  conductor  the  radiation  resistance  of 
the  antenna  or  array  is  increased  slightly  as  a 
result  of  the  increase  in  the  effective  diameter 
of  the  element.  Further,  if  we  split  just  one 


444  Antennas  and  Antenna  Matching 


THE  RADIO 


© 


h H 

-i  ^ _ HI  _ : zn}^ 


r — “ — -----  ---  ^ 


"TWIN-LEAD"  FLAT-TOP  OR 
CONDUCTORS  SPACED  1 " TO  T2" 


f»-  "TWIN-LEAD" 


300  OHMS 
ANY  LENGTH 


© E 


300-600 
OHM  FEEDERS 


Figure  30 

THE  GAMMA  MATCH  FOR  CONNECTING 
AN  UNBALANCED  COAXIAL  LINE  TO  A 
BALANCED  DRIVEN  ELEMENT 


300 -«00 
OHM  FEEDERS 


Figure  29 

FOLDED-ELEMENT  MATCHING  SYSTEMS 

Drawing  (A)  above  shows  a half-wave  made 
up  to  two  parallel  wires.  If  one  of  the  wires 
Is  broken  as  In  (B)  and  the  feeder  cannected. 
the  feed-point  Impedance  Is  multiplied  by 
four;  such  an  antenna  Is  commanly  called  a 
^'folded  doublet."  The  feed-point  Impedance 
for  a simple  half-wave  doublet  fed  In  this 
manner  Is  approximately  300  ohms,  depend- 
ing upon  antenna  height.  Drawing  (C)  shows 
how  the  feed-point  impedance  can  be  multi- 
plied by  a factor  greater  than  four  by  making 
the  half  of  the  element  that  is  broken  smaller 
in  diameter  than  the  unbroken  half.  An  ex- 
tension of  the  principles  of  (B)  and  (C)  is 
the  arrangement  shown  at  (D)  where  the  sec- 
tion into  which  the  feeders  are  connected  is 
considerably  shorter  than  the  driven  element. 
This  system  Is  most  convenient  when  the 
driven  element  is  too  long  (such  as  for  a 
28-Mc.  or  14-Mc.  array)  for  a convenient 
mechanical  arrangement  of  the  system  shown 
at(C). 


wire  of  such  a radiator,  as  shown  in  figure  29, 
the  effective  feed-point  resistance  of  the  an- 
tenna or  array  will  be  increased  by  a factor 
of  where  N is  equal  to  the  number  of  con- 
ductors, all  in  parallel,  of  the  same  diameter 
in  the  array.  Thus  if  there  are  two  conductors 
of  the  same  diametet  in  the  driven  element  ot 
the  antenna  the  feed-point  resistance  will  be 
multiplied  by  2^  or  4.  If  the  antenna  has  a ra- 
diation resistance  of  75  ohms  its  feed-point 
resistance  will  be  300  ohms,  this  is  the  case 


of  the  conventional  folded-dipole  as  shown  in 
figure  29B. 

If  three  wires  are  used  in  the  driven  radia- 
tor the  feed-point  resistance  is  increased  by 
a factor  of  9;  if  four  wires  are  used  the  im- 
pedance is  increased  by  a factor  of  16,  and 
so  on.  In  certain  cases  when  feeding  a para- 
sitic artay  it  is  desitable  to  have  an  imped- 
ance step  up  different  from  the  value  of  4:1 
obtained  with  two  elements  of  the  same  dia- 
meter and  9:1  with  three  elements  of  the  same 
diametet.  Intermediate  values  of  impedance 
step  up  may  be  obtained  by  using  two  elements 
of  different  diameter  for  the  complete  driven 
element  as  shown  in  figute  29C.  If  the  con- 
ductor that  is  broken  for  the  feeder  is  of  small- 
er diameter  than  the  other  conductor  of  the 
radiator,  the  impedance  step  up  will  be  greater 
than  4:1.  On  the  other  hand  if  the  larger  of  the 
two  elements  is  btoken  for  the  feeder  the  im- 
pedance step  up  will  be  less  than  4:1. 

The  "T”  Match  A method  of  matching  a bal- 
anced low-impedance  trans- 
mission line  to  the  driven  element  of  a para- 
sitic array  is  the  T match  illustrated  in  figute 
29D.  This  method  is  an  adaptation  of  the 
multi-wire  doublet  principle  which  is  more 
practicable  for  lower-frequency  parasitic  ar- 
rays such  as  those  for  use  on  the  14-Mc.  and 
28-Mc.  bands.  In  the  system  a section  of  tub- 
ing of  ^proximately  one-half  the  diameter  of 
the  driven  element  is  spaced  about  four  in- 
ches below  the  driven  element  by  means  of 
clamps  which  hold  the  T- section  mechanically 
and  which  make  electrical  connection  to  the 
driven  element.  The  length  of  the  T-section 
is  normally  between  15  and  30  inches  each 
side  of  the  center  of  the  dipole  for  transmis- 
sion lines  of  300  to  600  ohms  impedance,  as- 
suming 28-Mc.  operation.  In  series  with  each 
leg  of  the  T-section  and  the  ttansmission  line 
is  a series  resonating  capacitor.  These  two 
capacitors  tune  out  the  teactance  of  the  T- 


HANDBOOK 


Matching  Systems 


445 


section.  If  they  are  not  used,  the  T-section 
will  detune  the  dipole  when  the  T-section  is 
attached  to  it.  The  two  capacitors  may  be 
ganged  together,  and  once  adjusted  for  mini- 
mum detuning  action,  they  may  be  locked.  A 
suitable  housing  should  be  devised  to  protect 
these  capacitors  from  the  weather.  Additional 
information  on  the  adjustment  of  the  T-match 
is  given  in  the  chapter  covering  rotary  beam 
antennas. 

The  "Gammo"  Match  An  unbalanced  version  of 
the  T-match  may  be  used 
to  feed  a dipole  from  an  unbalanced  coaxial 
line.  Such  a device  is  called  a Gamma  Match, 
and  is  illustrated  in  figure  30. 

The  length  of  the  Gamma  rod  and  the  spac- 
ing of  it  from  the  dipole  determine  the  imped- 
ance level  at  the  transmission  line  end  of  the 
rod.  The  series  capacitor  is  used  to  tune  out 
the  reactance  introduced  into  the  system  by 
the  Gamma  tod.  The  adjustment  of  the  Gamma 
Match  is  discussed  in  the  chapter  covering 
rotary  beam  antennas. 

Matching  Stubs  By  connecting  a resonant 
section  of  transmission  line 
(called  a matching  stub)  to  either  a voltage  or 
current  loop  and  attaching  parallel-wire  non- 
resonant feeders  to  the  resonant  stub  at  a 
suitable  voltage  (impedance)  point,  standing 
waves  on  the  line  may  be  virtually  eliminated. 
The  stub  is  made  to  serve  as  an  auto-trans- 
former. Stubs  are  particularly  adapted  to 
matching  an  open  line  to  certain  directional 
arrays,  as  will  be  described  later. 

Voltage  Feed  When  the  stub  attaches  to  the 

antenna  at  a voltage  loop,  the 
stub  should  be  a quarter  wavelength  long 
electrically,  and  be  shorted  at  the  bottom  end. 
The  stub  can  be  resonated  by  sliding  the 
shotting  bar  up  and  down  before  the  non-teso- 
nant  feeders  are  attached  to  the  stub,  the  an- 
tenna being  shock-excited  from  a separate 
radiator  during  the  process.  Slight  errors  in 
the  length  of  the  radiator  can  be  compensated 
for  by  adjustment  of  the  stub  if  both  sides  of 
the  stub  are  connected  to  the  radiator  in  a 
symmetrical  manner.  Where  only  one  side  of 
the  stub  connects  to  the  radiating  system,  as 
in  the  Zepp  and  in  certain  antenna  arrays,  the 
radiator  length  must  be  exactly  right  in  order 
to  prevent  excessive  unbalance  in  the  untuned 
line. 

A dial  lamp  may  be  placed  in  the  center  of 
the  shorting  stub  to  act  as  an  r-f  indicator. 

Current  Feed  When  a stub  is  used  to  current- 

feed  a radiator,  the  stub  should 
either  be  left  open  at  the  bottom  end  instead 
of  shorted,  or  else  made  a half  wave  long. 


ANTENNA 


Figure  31 

MATCHING-STUB  APPLICATIONS 

An  end-fed  half-wave  antenna  with  a quarter- 
wave  shorted  stub  is  shown  at  (A).  (B)  shows 
the  use  of  a half-wave  shorted  stub  to  feed 
a relatively  low  impedance  point  such  as  the 
center  of  the  driven  element  of  a parasitic 
array,  or  the  center  of  a half-wave  dipole. 
The  use  of  on  open-ended  quarter-wave  stub 
to  feed  a low  impedance  is  illustrated  at 
(C),  (D)  shows  the  conventional  use  of  a 
shorted  quarter-wave  stub  to  voltage  feed 
two  half-wave  antennas  with  a 180°  phase 
difference. 


446  Antennas  and  Antenna  Matching 


THE  RADIO 


The  open  stub  should  be  resonated  in  the 

same  manner  as  the  shorted  stub  before  at- 
taching the  transmission  line;  however,  in 
this  case,  it  is  necessary  to  prune  the  stub 
to  resonance,  as  there  is  no  shorting  bat. 

Sometimes  it  is  handy  to  have  a stub  hang 
from  the  radiator  to  a point  that  can  be  reached 
from  the  ground,  in  order  to  facilitate  adjust- 
ment of  the  position  of  the  transmission-line 
attachment.  For  this  reason,  a quarter-wave 
stub  is  sometimes  made  three-quarters  wave- 
length long  at  the  higher  frequencies,  in  order 
to  bring  the  bottom  nearer  the  ground.  Opera- 
tion with  any  odd  number  of  quarter  waves  is 
the  same  as  for  a quarter- wave  stub. 

Any  number  of  half  waves  can  be  added  to 
either  a quarter-wave  stub  or  a half-wave  stub 
without  disturbing  the  operation,  though  losses 
and  frequency  sensitivity  will  be  lowest  if 
the  shortest  usable  stub  is  employed.  See  fig- 
ure 31. 


Stub  Length 

Current- Fed 

Voltage-Fed 

(Eiectrical ) 

Radiator 

Radiator 

'/4-ete. 

Open 

Shorted 

wavelengths 

Stub 

Stub 

H-l-IH-2-etc. 

Shorted 

Open 

wavelengths 

Stub 

Stub 

Linear  R-F  A resonant  quarter-wave  line 
Transformers  has  the  unusual  property  of 
acting  much  as  a transformer. 
Let  us  take,  for  example,  a section  consisting 
of  no.  12  wire  spaced  6 inches,  which  happens 
to  have  a surge  impedance  of  600  ohms.  Let 
the  far  end  be  terminated  with  a pure  resist- 
ance, and  let  the  neat  end  be  fed  with  radio- 
frequency energy  at  the  frequency  for  which 
the  line  is  a quarter  wavelength  long.  If  an 
impedance  measuring  set  is  used  to  measure 
the  impedance  at  the  near  end  while  the  im- 
pedance at  the  fat  end  is  varied,  an  interest- 
ing relationship  between  the  600-ohm  charac- 
teristic surge  impedance  of  this  particular 
quarter-wave  matching  line,  and  the  imped- 
ance at  the  ends  will  be  discovered. 

When  the  impedance  at  the  far  end  of  the 
line  is  the  same  as  the  characteristic  surge 
impedance  of  the  line  itself  (600  ohms),  the 
impedance  measured  at  the  near  end  of  the 
quarter-wave  line  will  also  be  found  to  be 
600  ohms. 

Under  these  conditions,  the  line  would  not 
have  any  standing  waves  on  it,  since  it  is 
terminated  in  its  characteristic  impedance. 
Now,  let  the  resistance  at  the  far  end  of  the 
line  be  doubled,  or  changed  to  1200  ohms. 
The  impedance  measured  at  the  near  end  of 
the  line  will  be  found  to  have  been  cut  in 


half,  to  300  ohms.  If  the  resistance  at  the  far 

end  is  made  half  the  original  value  of  600 
ohms  or  300  ohms,  the  impedance  at  the  near 
end  doubles  the  original  value  of  600  ohms, 
and  becomes  1200  ohms.  As  one  resistance 
goes  up,  the  other  goes  down  proportionately. 

It  will  always  be  found  that  the  character- 
istic surge  impedance  of  the  quarter-wave 
matching  line  is  the  geometric  mean  between 
the  impedance  at  both  ends.  This  relationship 
is  shown  by  the  following  formula: 

Zms  — 'J  Za  Zl 

where 

Zys  = Impedance  of  matching  section. 

Za  = Antenna  resistance. 

Zl  = Line  impedance. 

Quarter-Wave  The  impedance  inverting  char- 

Matching  acteristic  of  a quarter-wave 

Transformers  section  of  transmission  line  is 

widely  used  by  making  such  a 
section  of  line  act  as  a quarter-wave  trans- 
former. The  Johnson  Q feed  system  is  a wide- 
ly known  application  of  the  quarter-wave  trans- 
former to  the  feeding  of  a dipole  antenna  and 
array  consisting  of  two  dipoles.  However,  the 
quarter-wave  transformer  may  be  used  in  a 
wide  number  of  applications  wherever  a trans- 
former is  required  to  match  two  impedances 
whose  geometric  mean  is  somewhere  between 
perhaps  25  and  750  ohms  when  transmission 
line  sections  can  be  used.  Paralleled  coaxial 
lines  may  be  used  to  obtain  the  lowest  im- 
pedance mentioned,  and  open-wire  lines  com- 
posed of  small  conductors  spaced  a moderate 
distance  may  be  used  to  obtain  the  higher  im- 
pedance. A short  list  of  impedances,  which 
may  be  matched  by  quarter-wave  sections  of 
transmission  line  having  specified  imped- 
ances, is  given  below. 


Load  or  Ant. 
Impedance  ^ 

300 

480 

600 

Feed-Line 

Impedance 

20 

98 

110 

Quarter- 

30 

95 

120 

134 

Wave 

50 

no 

139 

155 

Transformer 

75 

150 

190 

212 

Impedance 

100 

173 

220 

245 

Johnson- Q The  standard  form  of  Johnson- 

feed  System  Q feed  to  a doublet  is  shown 
in  figure  32.  An  impedance 
match  is  obtained  by  utilizing  a matching  sec- 
tion, the  surge  impedance  of  which  is  the  geo- 
metric mean  between  the  transmission  line 
surge  impedance  and  the  radiation  resistance 
of  the  radiator.  A sufficiently  good  match  usu- 


HANDBOOK 


Matching  Systems  447 


Lfee;t  = -§SS- 
F MC 


*' — J 

1 

TUBING -♦ 

Q MATCHING  SECTION 

L = 

234 

1 

•-ZQ=  Vzi  A Z2 

F(mc) 

J 

UNTUNED  LINE 
ANY  LENGTH 


Figure  32 

HALF-WAVE  RADIATOR  FED 
BY  "Q  BARS" 

The  Q matching  section  is  simply  a quarter- 
wave  transformer  whose  impedance  is  equal 
to  the  geometric  mean  between  the  imped- 
ance at  the  center  of  the  antenna  and  the 
Impedance  of  the  transmission  line  to  be 
used  to  feed  the  bottom  of  the  transformer- 
The  transformer  may  be  made  up  of  parallel 
tubing-  a four-wire  line-  or  any  other  type 
of  transmission  line  which  has  the  correct 
value  of  impedance. 


ally  can  be  obtained  by  either  designing  or 
adjusting  the  matching  section  for  a dipole  to 
have  a surge  impedance  that  is  the  geometric 
mean  between  the  line  impedance  and  72  ohms, 
the  latter  being  the  theoretical  radiation  re- 
sistance of  a half-wave  doublet  either  infi- 
nitely high  or  a half  wave  above  a perfect 
ground. 

Though  the  radiation  resistance  may  depart 
somewhat  from  72  ohms  under  actual  condi- 
tions, satisfactory  results  will  be  obtained 
with  this  assumed  value,  so  long  as  the  dipole 
radiator  is  more  than  a quarter  wave  above 
effective  earth,  and  reasonably  in  the  clear. 
The  small  degree  of  standing  waves  intro- 
duced by  a slight  mismatch  will  not  increase 
the  line  losses  appreciably,  and  any  small 
amount  of  reactance  present  can  be  tuned  out 
at  the  transmitter  termination  with  no  bad  ef- 
fects. If  the  reactance  is  objectionable,  it  may 
be  minimized  by  making  the  untuned  line  an 
integral  number  of  quarter  waves  long. 

A Q-matched  system  can  be  adjusted  pre- 
cisely, if  desired,  by  constructing  a matching 
section  to  the  calculated  dimensions  with  pro- 
vision for  varying  the  spacing  of  the  Q sec- 
tion conductors  slightly,  after  the  untuned 
line  has  been  checked  for  standing  waves. 


Center  to 
Center 
Spacing 
in  Inches 

Impedance 
in  Ohms 
for 

Diameters 

Impedance 
in  Ohms 
for  Vi  " 
Diameters 

1 

170 

250 

1.25 

188 

277 

1.5 

207 

298 

1.75 

225 

318 

2 

248 

335 

PARALLEL  TUBING  SURGE  IMPEDANCE  FOR 
MATCHING  SECTIONS 


The  Collins  The  advantage  of  unbal- 

Tronsmission  Line  anced  output  networks 

Matching  System  for  transmitters  are  nu- 

merous; however  this  out- 
put system  becomes  awkward  when  it  is  de- 
sired to  feed  an  antenna  system  utilizing  a 
balanced  input.  For  some  time  the  Collins 
Radio  Co.  has  been  experimenting  with  a bal- 
un  and  tapered  line  system  for  matching  a co- 
axial output  transmitter  to  an  open-wire  bal- 
anced transmission  line.  Considerable  success 
has  been  obtained  and  matching  systems  good 
over  a frequency  range  as  great  as  four  to  one 
have  been  developed.  Illustrated  in  figure  33 
is  one  type  of  matching  system  which  is  prov- 
ing satisfactory  over  this  range.  Z,  is  the 
transmitter  end  of  the  system  and  may  be  any 
length  of  52-ohm  coaxial  cable.  Zj  is  one- 
quarter  wavelength  long  at  the  mid- frequency 
of  the  range  to  be  covered  and  is  made  of  75 
ohm  coaxial  cable.  Za  is  a quarter-wavelength 
shorted  section  of  cable  at  the  mid-frequency. 
^0  3ttd  Zj)  forms  a 200-ohm  quarter-wave 
section.  The  Za  section  is  formed  of  a con- 
ductor of  the  same  diameter  as  Zj.  The  differ- 
ence in  length  between  Za  and  Z,  is  accounted 
for  by  the  fact  that  Zj  is  a coaxial  conductor 
with  a solid  dielectric,  whereas  the  dielectric 
for  Zo  is  air.  Zj  is  one-quarter  wavelength 
long  at  the  mid-frequency  and  has  an  imped- 


Figure  33 

COLLINS  TRANSMISSION  LINE  MATCHING 
SYSTEM 


A wide-bond  system  for  matching  a 52-ohm 
coaxial  line  to  a balanced  300-ohm  line  over 
a 4:1  wide  frequency  range. 


44  8 Antennas  and  Antenna  Matching 


THE  RADIO 


ance  of  123  ohms.  Z^  is  one-quarter  wavelength 
long  at  the  mid-frequency  and  has  an  imped- 
ance of  224  ohms.  Z;  is  the  balanced  line  to 
be  matched  (in  this  case  300  ohms)  and  may  be 
any  length. 

Other  system  parameters  for  different  output 
and  input  impedances  may  be  calculated  from 
the  following: 

Transformation  ratio  (r)  for  each  section  is: 


Where  N is  the  number  of  sections.  In  the 
above  case, 


Impedance  between  sections,  as  Zj-j,  is  r 
times  the  preceding  section.  Zj-j  = r x Z„  and 
Z3-4  = r X Z2-3. 

Mid-frequency  (m): 

F.  +F2 

m = 

2 

7 + 30 

For  40-20-10  meters  = = 18.5  Me. 

2 

and  one-quarter  wavelength  = 12  feet. 

For  20-10-6  meters  - ^ 

2 

and  one-quarter  wavelength  =5.5  feet. 

The  impedances  of  the  sections  are: 

Zj  = V Zj  X Z2.3 
Z3  — yf  Z2-3  X Z3.4 
Z,  = V Z3.,  X Z, 

Generally,  the  larger  number  of  taper  sec- 
tions the  greater  will  be  the  bandwidth  of  the 
system. 

22-9  Antenna 

Construction 

The  foregoing  portion  of  this  chapter  has 


been  concerned  primarily  with  the  electrical 
characteristics  and  considerations  of  anten- 
nas. Some  of  the  physical  aspects  and  mech- 
anical problems  incident  to  the  actual  erec- 
tion of  antennas  and  arrays  will  be  discussed 
in  the  following  section. 

Up  to  60  feet,  there  is  little  point  in  using 
mast-type  antenna  supports  unless  guy  wires 
either  must  be  eliminated  or  kept  to  a mini- 
mum. While  a little  more  difficult  to  erect,  be- 
cause of  their  floppy  nature,  fabricated  wood 
poles  of  the  type  to  be  described  will  be  just 
as  satisfactory  as  more  rigid  types,  provided 
many  guy  wires  are  used. 

Rather  expensive  when  purchased  through 
the  regular  channels,  40-  and  50-foot  tele- 
phone poles  sometimes  can  be  obtained  quite 
reasonably.  In  the  latter  case,  they  are  hard 
to  beat,  inasmuch  as  they  require  no  guying 
if  set  in  the  ground  six  feet  (standard  depth), 
and  the  resultant  pull  in  any  lateral  direction 
is  not  in  excess  of  a hundred  pounds  or  so. 

For  heights  of  80  to  100  feet,  either  three- 
sided  or  four-sided  lattice  type  masts  are  most 
practicable.  They  can  be  made  self-support- 
ing, but  a few  guys  will  enable  one  to  use  a 
smaller  cross  section  without  danger  from 
high  winds.  The  torque  exerted  on  the  base 
of  a high  self-supporting  mast  is  terrific  dur- 
ing a strong  wind. 

The  **A-Frome*'  Figures  34A  and  34B  show 
Most  the  standard  method  of  con- 

struction of  the  A-frame 
type  of  mast.  This  type  of  mast  is  quite  fre- 
quently used  since  there  is  only  a moderate 
amount  of  work  involved  in  the  construction 
of  the  assembly  and  since  the  material  cost  is 
relatively  small.  The  three  pieces  of  selected 
2 by  2 are  first  set  up  on  three  sawhorses  or 
boxes  and  the  holes  drilled  for  the  three  Va- 
inch  bolts  through  the  center  of  the  assembly. 
Then  the  base  legs  are  spread  out  to  about  6 
feet  and  the  bottom  braces  installed.  Then  the 
upper  braces  and  the  cross  pieces  are  in- 
stalled and  the  assembly  given  several  coats 
of  good-quality  paint  as  a protection  against 
weathering. 

Figure  34C  shows  another  common  type  of 
mast  which  is  made  up  of  sections  of  2 by  4 
placed  end-to-end  with  stiffening  sections  of 
1 by  6 bolted  to  the  edge  of  the  2 by  4 sec- 
tions. Both  types  of  masts  will  require  a set 
of  top  guys  and  another  set  of  guys  about  one- 
third  of  the  way  down  from  the  top.  Two  guys 
spaced  about  90  to  100  degrees  and  pulling 
against  the  load  of  the  antenna  will  normally 
be  adequate  for  the  top  guys.  Three  guys  are 
usually  used  at  the  lower  level,  with  one  di- 
rectly behind  the  load  of  the  antenna  and  two 
more  spaced  120  degrees  from  the  rear  guy. 

The  raising  of  the  mast  is  made  much  easier 


HANDBOOK 


Antenna  Construction  449 


Figure  34 

TWO  SIMPLE  WOOD  MASTS 

Shown  of  (A)  is  the  method  of  as- 
sembly, and  at  (B)  is  the  completed 
structure,  of  the  conventional  "A- 
frame"  antenna  mast.  At  (C)  is 
shown  a structure  which  is  heavier 
but  more  stable  than  the  A-frame 
for  heights  above  about  40  feet. 


if  a gin  pole  about  20  feet  high  is  installed 
about  30  or  40  feet  to  the  tear  of  the  direction 
in  which  the  antenna  is  to  be  raised.  A line 
from  a pulley  on  the  top  of  the  gin  pole  is  then 
tun  to  the  top  of  the  pole  to  be  raised.  The 
gin  pole  comes  into  play  when  the  center  of 
the  mast  has  been  raised  10  to  20  feet  above 
the  ground  and  an  additional  elevated  pull  is 
requited  to  keep  the  top  of  the  mast  coming 
up  as  the  center  is  raised  further  above 
ground. 

Using  TV  Masts  Steel  tubing  masts  of  the 
telescoping  variety  are  wide- 
ly available  at  a moderate  price  for  use  in  sup- 
porting television  antenna  arrays.  These  masts 
usually  consist  of  several  10-foot  lengths  of 
electrical  metal  tubing  (EMT)  of  sizes  such 
that  the  sections  will  telescope.  The  30-foot 
and  40-foot  lengths  are  well  suited  as  masts 
for  supporting  antennas  and  arrays  of  the 
types  used  on  the  amateur  bands.  The  masts 
are  constructed  in  such  a manner  that  the  bot- 
tom 10-foot  length  may  be  guyed  permanently 
before  the  other  sections  are  raised.  Then  the 
upper  sections  may  be  extended,  beginning 
with  the  top-mast  section,  until  the  mast  is  at 
full  length  (provided  a strong  wind  is  not  blow- 
ing) following  which  all  the  guys  may  be  an- 
chored. It  is  important  that  there  be  no  load 
on  the  top  of  the  mast  when  the  "vertical” 
raising  method  is  to  be  employed. 

Guy  Wires  Guy  wires  should  never  be  pulled 
taut;  a small  amount  of  slack  is 
desirable.  Galvanized  wire,  somewhat  heavier 


than  seems  sufficient  for  the  job,  should  be 
used.  The  heavier  wire  is  a little  harder  to 
handle,  but  costs  only  a little  more  and  takes 
longer  to  rust  through.  Care  should  be  taken 
to  make  sure  that  no  kinks  exist  when  the  pole 
or  tower  is  ready  for  erection,  as  the  wire  will 
be  greatly  weakened  at  such  points  if  a kink 
is  pulled  tight,  even  if  it  is  later  straightened. 

If  "dead  men”  ate  used  for  the  guy  wire 
terminations,  the  wire  or  rod  reaching  from  the 
dead  men  to  the  surface  should  be  of  non-rust- 
ing material,  such  as  brass,  or  given  a heavy 
coating  of  asphalt  or  other  protective  sub- 
stance to  prevent  destructive  action  by  the 
damp  soil.  Galvanized  iron  wire  will  last  only 
a short  time  when  buried  in  moist  soil. 

Only  strain-type  (compression)  insulators 
should  be  used  for  guy  wires.  Regular  ones 
might  be  sufficiently  strong  for  the  job,  but  it 
is  not  worth  taking  chances,  and  egg-type 
strain  halyard  insulators  are  no  more  ex- 
pensive. 

Only  a brass  or  bronze  pulley  should  be 
used  for  the  halyard,  as  a high  pole  with  a 
rusted  pulley  is  truly  a sad  affair.  The  bear- 
ing of  the  pulley  should  be  given  a few  drops 
of  heavy  machine  oil  before  the  pole  or  tower 
is  raised.  The  halyard  itself  should  be  of  good 
material,  preferably  water-proofed.  Hemp  rope 
of  good  quality  is  better  than  window  sash 
cord  from  several  standpoints,  and  is  less  ex- 
pensive. Soaking  it  thoroughly  in  engine  oil  of 
medium  viscosity,  and  then  wiping  it  off  with 
a tag,  will  not  only  extend  its  life  but  minim- 
ize shrinkage  in  wet  weather.  Because  of  the 
difficulty  of  replacing  a broken  halyard  it  is 
a good  idea  to  replace  it  periodically,  without 


450  Antennas  and  Antenna  Matching 


THE  RADIO 


waiting  for  it  to  show  excessive  wear  or  de- 
terioration. 

It  is  an  excellent  idea  to  tie  both  ends  of 
the  halyard  line  together  in  the  manner  of  a 
flag-pole  line.  Then  the  antenna  is  tied  onto 
the  place  where  the  two  ends  of  the  halyard 
are  joined.  This  procedure  of  making  the  hal- 
yard into  a loop  prevents  losing  the  top  end 
of  the  halyard  should  the  antenna  break  near 
the  end,  and  it  also  prevents  losing  the  hal- 
yard completely  should  the  end  of  the  halyard 
carelessly  be  allowed  to  go  free  and  be  pulled 
through  the  pulley  at  the  top  of  the  mast  by 
the  antenna  load.  A somewhat  longer  piece 
of  line  is  required  but  the  insurance  is  well 
worth  the  cost  of  the  additional  length  of  rope. 

Trees  as  Often  a tall  tree  can  be  called  up- 
Supports  on  to  support  one  end  of  an  anten- 
na, but  one  should  not  attempt  to 
attach  anything  to  the  top,  as  the  swaying  of 
the  top  of  the  tree  during  a heavy  wind  will 
complicate  matters. 

If  a tree  is  utilized  for  support,  provision 
should  be  made  for  keeping  the  antenna  taut 
without  submitting  it  to  the  possibility  of  be- 
ing severed  during  a heavy  wind.  This  can  be 
done  by  the  simple  expedient  of  using  a pul- 
ley and  halyard,  with  weights  attached  to  the 
lower  end  of  the  halyard  to  keep  the  antenna 
taut.  Only  enough  weight  to  avoid  excessive 
sag  in  the  antenna  should  be  tied  to  the  hal- 
yard, as  the  continual  swaying  of  the  tree  sub- 
mits the  pulley  and  halyard  to  considerable 
wear. 

Galvanized  iron  pipe,  or  steel-tube  conduit, 
is  often  used  as  a vertical  radiator,  and  is 
quite  satisfactory  for  the  purpose.  However, 
when  used  for  supporting  antennas,  it  should 
be  remembered  that  the  grounded  supporting 
poles  will  distort  the  field  pattern  of  a verti- 
cally polarized  antenna  unless  spaced  some 
distance  from  the  radiating  portion. 

Painting  The  life  of  a wood  mast  or  pole  can 
be  increased  several  hundred  per 
cent  by  protecting  it  from  the  elements  with  a 
coat  or  two  of  paint.  And,  of  course,  the  ap- 
pearance is  greatly  enhanced.  The  wood  should 
first  be  given  a primer  coat  of  flat  white  out- 
side house  paint,  which  can  be  thinned  down 
a bit  to  advantage  with  second-grade  linseed 
oil.  For  the  second  coat,  which  should  not  be 
applied  until  the  first  is  thoroughly  dry,  alumi- 
num paint  is  not  only  the  best  from  a preserva- 
tive standpoint,  but  looks  very  well.  This  type 
of  paint,  when  purchased  in  quantities,  is  con- 
siderably cheaper  than  might  be  gathered  from 
the  price  asked  for  quarter-pint  cans. 

Portions  of  posts  or  poles  below  the  surface 
of  the  soil  can  be  protected  from  termites  and 
moisture  by  painting  with  creosote.  While  not 


so  strong  initially,  redwood  will  deteriorate 
much  more  slowly  when  buried  than  will  the 
white  woods,  such  as  pine. 

Antenna  Wire  The  antenna  or  array  itself 
presents  no  especial  problem. 
A few  considerations  should  be  borne  in  mind, 
however.  For  instance,  soft-drawn  copper 
should  not  be  used,  as  even  a short  span  will 
stretch  several  per  cent  after  whipping  around 
in  the  wind  a few  weeks,  thus  affecting  the 
resonant  frequency.  Enameled-copper  wire, 
as  ordinarily  available  at  radio  stores,  is  us- 
ally  soft  drawn,  but  by  tying  one  end  to  some 
object  such  as  a telephone  pole  and  the  other 
to  the  frame  of  an  auto,  a few  husky  tugs  can 
be  given  and  the  wire,  after  stretching  a bit, 
is  equivalent  to  hard  drawn. 

Where  a long  span  of  wire  is  required,  or 
where  heavy  insulators  in  the  center  of  the 
span  result  in  considerable  tension,  copper- 
clad  steel  wire  is  somewhat  better  than  hard- 
drawn  copper.  It  is  a bit  mote  expensive, 
though  the  cost  is  far  from  prohibitive.  The 
use  of  such  wire,  in  conjunction  with  strain 
insulators,  is  advisable,  where  the  antenna 
would  endanger  persons  or  property  should  it 
break. 

For  transmission  lines  and  tuning  stubs 
steel-core  or  hard-drawn  wire  will  prove  awk- 
ward to  handle,  and  soft-drawn  copper  should, 
therefore,  be  used.  If  the  line  is  long,  the 
strain  can  be  eased  by  supporting  it  at  several 
points. 

More  important  from  an  electrical  standpoint 
than  the  actual  size  of  wire  used  is  the  sol- 
dering of  joints,  especially  at  current  loops 
in  an  antenna  of  low  radiation  resistance.  In 
fact,  it  is  good  practice  to  solder  all  joints, 
thus  insuring  quiet  operation  when  the  anten- 
na is  used  for  receiving. 

Insulation  A question  that  often  arises  is 
that  of  insulation.  It  depends,  of 
course,  upon  the  r-f  voltage  at  the  point  at 
which  the  insulator  is  placed.  The  r-f  voltage, 
in  turn,  depends  upon  the  distance  from  a cur- 
rent node,  and  the  radiation  resistance  of  the 
antenna.  Radiators  having  low  radiation  re- 
sistance have  very  high  voltage  at  the  voltage 
loops;  consequently,  better  than  usual  insula- 
tion is  advisable  at  those  points. 

Open-wire  lines  operated  as  non-resonant 
lines  have  little  voltage  across  them;  hence 
the  most  inexpensive  ceramic  types  are  suffi- 
ciently good  electrically.  With  tuned  lines,  the 
voltage  depends  upon  the  amplitude  of  the 
standing  waves.  If  they  are  very  great,  the 
voltage  will  reach  high  values  at  the  voltage 
loops,  and  the  best  spacers  available  are  none 
too  good.  At  the  current  loops  the  voltage  is 
quite  low,  and  almost  anything  will  suffice. 


HANDBOOK 


Antenna  Coupling  Systems  451 


When  insulators  are  subject  to  very  high  r-f 
voltages,  they  should  be  cleaned  occasionally 
if  in  the  vicinity  of  sea  water  or  smoke.  Salt 
scum  and  soot  are  not  readily  dislodged  by 
rain,  and  when  the  coating  becomes  heavy 
enough,  the  efficiency  of  the  insulators  is 
greatly  impaired. 

If  a very  pretentious  installation  is  to  be 
made,  it  is  wise  to  check  up  on  both  under- 
writer's rules  and  local  ordinances  which 
might  be  applicable.  If  you  live  anywhere  near 
an  airport,  and  are  contemplating  a tall  pole, 
it  is  best  to  investigate  possible  regulations 
and  ordinances  pertaining  to  towers  in  the  dis- 
trict, before  starting  construction. 


22-10  Coupling  to  the 

Antenna  System 

When  coupling  an  antenna  feed  system  to  a 
transmitter  the  most  important  considerations 
are  as  follows:  (1)  means  should  be  provided 
for  varying  the  load  on  the  amplifier;  (2)  the 
two  tubes  in  a push-pull  amplifier  should  be 
equally  loaded;  (3)  the  load  presented  to  the 
final  amplifier  should  be  resistive  (non-reac- 
tive) in  character;  and  (4)  means  should  be 
provided  to  reduce  harmonic  coupling  between 
the  final  amplifier  plate  tank  circuit  and  the 
antenna  or  antenna  transmission  line  to  an  ex- 
tremely low  value. 

The  Transmitter-  The  problem  of  coupling  the 
Loading  Problem  power  output  of  a high-fre- 
quency or  v-h-f  transmitter 
to  the  radiating  portion  of  the  antenna  system 
has  been  materially  complicated  by  the  virtual 
necessity  for  eliminating  interference  to  TV  re- 
ception. However,  the  TVI- elimination  portion 
of  the  problem  may  always  be  accomplished 
by  adequate  shielding  of  the  transmitter,  by 
filtering  of  the  control  and  power  leads  which 
enter  the  transmitter  enclosure,  and  by  the  in- 
clusion of  a harmonic-attenuating  filter  be- 


tween the  output  of  the  transmitter  and  the  an- 
tenna system. 

Although  TVI  may  be  eliminated  through  in- 
clusion of  a filter  between  the  output  of  a 
shielded  transmitter  and  the  antenna  system, 
the  fact  that  such  a filter  must  be  included  in 
the  link  between  transmitter  and  antenna  makes 
it  necessary  that  the  transmitter-loading  prob- 
lem be  re-evaluated  in  terms  of  the  necessity 
for  inclusion  of  such  a filter. 

Harmonic-attenuating  filters  must  be  oper- 
ated at  an  impedance  level  which  is  close  to 
their  design  value;  therefore  they  must  operate 
into  a resistive  termination  substantially  equal 
to  the  characteristic  impedance  of  the  filter. 
If  such  filters  are  operated  into  an  impedance 
which  is  not  resistive  and  approximately  equal 
to  their  characteristic  impedance:  (1)  the  ca- 
pacitors used  in  the  filter  sections  will  be 
subjected  to  high  peak  voltages  and  may  be 
damaged,  (2)  the  harmonic-attenuating  proper- 
ties of  the  filter  will  be  decreased,  and  (3)  the 
impedance  at  the  input  end  of  the  filter  will 
be  different  from  that  seen  by  the  filter  at  the 
load  end  (except  in  the  case  of  the  half-wave 
type  of  filter).  It  is  therefore  important  that 
the  filter  be  included  in  the  transmitter-to-an- 
tenna  circuit  at  a point  where  the  impedance 
is  close  to  the  nominal  value  of  the  filter,  and 
at  a point  where  this  impedance  is  likely  to 
remain  fairly  constant  with  variations  in  fre- 
quency. 

Block  Diagrams  of  There  are  two  basic 

Tronsmittor-to-Antenna  arrangements  which 

Coupling  Systems  include  all  the  provi- 

sions required  In  the 
transmittet-to-antenna  coupling  system,  and 
which  permit  the  harmonic-attenuating  filter  to 
be  placed  at  a position  in  the  coupling  system 
where  it  can  be  operated  at  an  impedance 
level  close  to  its  nominal  value.  These  ar- 
rangements are  illustrated  in  block-diagram 
form  in  figures  35  and  36. 

The  arrangement  of  figure  35  is  recommend- 
ed for  use  with  a single-band  antenna  system, 
such  as  a dipole  or  a rotatable  array,  wherein 
an  impedance  matching  system  is  included 


AHIELD ^ AT  TRANSMITTER  AT  ANTENNA 


EXCITER 

1 

^impedance!  loAn.ariM/- 

PORTION 

h 

AMPLIFIER  ” 

OJU5TMENTp-r  I TRANSMISSION 

Ml  MATCHING 

^T  ANTENNA^  srsrtM 

Figure  35 

ANTENNA  COUPLING  SYSTEM 

The  harmonic  suppressing  antenna  coupling  system  illustrated  above  is  for  use  when  the  anten- 
no  transmission  tine  has  a tow  standing-wave  ration  and  when  the  characteristic  impedance  of 
the  antenna  transmission  line  is  the  some  as  the  nominal  impedance  of  the  low-pass  harmonic- 

attenuating  filter. 


452  Antennas  and  Antenna  Matching 


THE  RADIO 


TRANSMITTER  AT  ANTENNA 


COUPLING  1 ANTENNA  L J'MJx^uVSILJrADIATING 

PORTION 

AMPLIFIER 

ADJUSTMENTj  , | sySTEM  | | | TRANSMISSION  |ATANTeNNA|  \ 

Figure  36 

ANTENNA  COUPLING  SYSTEM 


The  antenna  coupling  system  illustrated  above  is  for  use  when  the  antenna  transmission  line 
does  not  have  the  same  characteristic  impedance  as  the  TVI  filter^  and  when  the  standing-wave 
ratio  on  the  antenna  transmission  line  may  or  may  not  be  low. 


within  or  adjacent  to  the  antenna.  The  feed 
line  coming  down  from  the  antenna  system 
should  have  a characteristic  impedance  equal 
to  the  nominal  impedance  of  the  harmonic  fil- 
ter, and  the  impedance  matching  at  the  anten- 
na should  be  such  that  the  standing-wave  ratio 
on  the  antenna  feed  line  is  less  than  2 to  1 
over  the  range  of  frequency  to  be  fed  to  the 
anterma.  Such  an  arrangement  may  be  used 
with  open-wire  line,  ribbon  or  tubular  line,  or 
with  coaxial  cable.  The  use  of  coaxial  cable 
is  to  be  recommended,  but  in  any  event  the 
impedance  of  the  antenna  transmission  line 
should  be  the  same  as  the  nominal  impedance 
of  the  harmonic  filter.  The  arrangement  of  fig- 
ure 35  is  more  or  less  standard  for  commercial- 
ly manufactured  equipment  for  amateur  and 
commercial  use  in  the  h-f  and  v-h-f  range. 

The  arrangement  of  figure  36  merely  adds 
an  antenna  coupler  between  the  output  of  the 
harmonic  attenuating  filter  and  the  antenna 
transmission  line.  The  antenna  coupler  will 
have  some  harmonic-attenuating  action,  but  its 
main  function  is  to  transform  the  impedance 
at  the  station  end  of  the  antenna  transmission 
line  to  the  nominal  value  of  the  harmonic  filter. 
Hence  the  arrangement  of  figure  36  is  more 
general  than  the  figure  35  system,  since  the 
inclusion  of  the  antenna  coupler  allows  the 
system  to  feed  an  antenna  transmission  line 
of  any  reasonable  impedance  value,  and  also 
without  regard  to  the  standing-wave  ratio 
which  might  exist  on  the  antenna  transmission 
line.  Antenna  couplers  are  discussed  in  a fol- 
lowing section. 

Output  Coupling  It  will  be  noticed  by  refer- 
Adjustment  ence  to  both  figure  35  and 

figure  36  that  a box  labeled 
Coupling  Adjustment  is  included  in  the  block 
diagram.  Such  an  element  is  necessary  in  the 
complete  system  to  afford  an  adjustment  in  the 
value  of  load  impedance  presented  to  the  tubes 
in  the  final  amplifier  stage  of  the  transmitter. 
The  impedance  at  the  input  terminal  of  the 
harmonic  filter  is  established  by  the  antenna, 
through  its  matching  system  and  the  antenna 


coupler,  if  used.  In  any  event  the  impedance 
at  the  input  terminal  of  the  harmonic  filter 
should  be  very  close  to  the  nominal  impedance 
of  the  filter.  Then  the  Coupling  Adjustment 
provides  means  for  transforming  this  imped- 
ance value  to  the  correct  operating  value  of 
load  impedance  which  should  be  presented  to 
the  final  amplifier  stage. 

There  are  two  common  ways  for  accomplish- 
ing the  antenna  coupling  adjustment,  as  illus- 
trated in  figures  37  and  38.  Figure  37  shows 
the  variable-link  arrangement  most  commonly 
used  in  home-constructed  equipment,  while  the 
pi-netowrk  coupling  arrangement  commonly 
used  in  commercial  equipment  is  illustrated  in 
figure  38.  Either  method  may  be  used,  and  each 
has  its  advantages. 

Variable-Link  The  variable-link  method  il- 
Coupling  lusttated  in  figure  37  has  the 

advantage  that  standard  man- 
ufactured components  may  be  used  with  no 
changes.  However,  for  greatest  bandwidth  of 
operation  of  the  coupling  circuit,  the  reactance 
of  the  link  coil,  L,  and  the  reactance  of  the 
link  tuning  capacitor,  C,  should  both  be  be- 
tween 3 and  4 times  the  nominal  load  imped- 
ance of  the  harmonic  filter.  This  is  to  say  that 
the  inductive  reactance  of  the  coupling  link  L 
should  be  tuned  out  or  resonated  by  capacitor 
C,  and  the  operating  Q of  the  L-C  link  circuit 
should  be  between  3 and  4.  If  the  link  coil  is 
not  variable  with  respect  to  the  tank  coil  of 
the  final  amplifier,  capacitor  C may  be  used 
as  a loading  control;  however,  this  system  is 
not  recommended  since  its  use  will  require 
adjustment  of  C whenever  a frequency  change 
is  made  at  the  transmitter.  If  L and  C are  made 
resonant  at  the  center  of  a band,  with  a link 
circuit  Q of  3 to  4,  and  coupling  adjustment  is 
made  by  physical  adjustment  of  L with  respect 
to  the  final  amplifier  tank  coil,  it  usually  will 
be  possible  to  operate  over  an  entire  amateur 
band  without  change  in  the  coupling  system. 
Capacitor  C normally  may  have  a low  voltage 
rating,  even  with  a high  power  transmitter,  due 
to  the  low  Q and  low  impedance  of  the  coupling 
circuit. 


HANDBOOK 


Pi-Network  Coupling  Systems  453 


TO  ANTENNA 
FEEDLINE  OR 
TO  ANTENNA 
COUPLER 


Capacitor  C should  be  adfusted  so  as  to  tuno  out  the  inductive  reactance  of  the  coupling  link, 
L.  Loading  of  the  amplifier  then  is  varied  by  physically  varying  the  coupling  between  the  plate 
tank  of  the  final  amplifier  and  the  antenna  coupling  link^ 


Pi-Network  The  pi-network  coupling  system 
Coupling  offers  two  advantages:  (1)  a me- 
chanical coupling  variation  is 
not  required  to  vary  the  loading  of  the  final  am- 
plifier, and  (2)  the  pi  network  (if  used  with  an 
operating  Q of  about  15)  offers  within  itself  a 
harmonic  attenuation  of  40  db  or  more,  in  ad- 
dition to  the  harmonic  attenuation  provided  by 
the  additional  harmonic  attenuating  filter-  Some 
commercial  equipments  (such  as  the  Collins 
amateur  transmitters)  incorporate  an  L network 
in  addition  to  the  pi  network,  for  accomplish- 
ing the  impedance  transformation  in  two  steps 
and  to  provide  additional  harmonic  attenuation. 

Tuning  the  Tuning  of  a pi-network 

Pi-Section  Coupler  coupling  circuit  such  as 
illustrated  in  figure  38  is 
accomplished  in  the  following  manner:  First 
remove  the  connection  between  the  output  of 
the  amplifier  and  the  harmonic  filter  (load). 
Tune  Cj  to  a capacitance  which  is  large  for 
the  band  in  use,  adding  suitable  additiontd  ca- 


pacitance by  switch  S if  operation  is  to  be  on 
one  of  the  lower  frequency  bands.  Apply  re- 
duced plate  voltage  to  the  stage  and  dip  to 
resonance  with  C,.  It  may  be  necessary  to  vary 
the  inductance  in  coil  L,  but  in  any  event  reso- 
nance should  be  reached  with  a setting  of  C, 
which  is  approximately  correct  for  the  desired 
value  of  operating  Q of  the  pi  network. 

Next,  couple  the  load  to  the  amplifier 
(through  the  harmonic  filter),  apply  reduced 
plate  voltage  again  and  dip  to  resonance  with 
C,.  If  the  plate  current  dip  with  load  is  too  low 
(taking  into  consideration  the  reduced  plate 
voltage),  decrease  the  capacitance  of  C,  and 
again  dip  to  resonance,  repeating  the  proce- 
dure until  the  correct  value  of  plate  current  is 
obtained  with  full  plate  voltage  on  the  stage. 
There  should  be  a relatively  small  change  re- 
quited in  the  setting  of  C,  (from  the  original 
setting  of  C,  without  load)  if  the  operating  Q 
of  the  network  is  correct  and  if  a large  value 
of  impedance  transformation  is  being  em- 
ployed—as  would  be  the  case  when  transform- 
ing from  the  plate  impedance  of  a single- 


TO  FEEDLINE 
OR  ANTENNA 
COUPLER 


The  design  of  pi-network  output  circuitsis  discussed  in  Chapter  Thirteen.  The  additional  output- 
end  shunting  capacitors  selected  by  switch  5 are  for  use  on  the  Tower  frequency  ranges.  Induc- 
tor L may  be  selected  by  a tap  switch,  it  may  be  continuously  variable,  or  plug-in  inductors 

may  2>e  used. 


454  Antennas  and  Antenna  Matching 


the  radio 


@ 


© 


Figure  39 

ALTERNATIVE  ANTENNA-COUPLER  CIRCUITS 

Plug-in  coils,  one  or  tv/o  variable  capacHors  of  fhe  split-stator  variety,  and  a system  of 
switches  or  plugs  and  jacks  may  be  used  in  the  antenna  coupler  to  accomplish  the  feeding  of 
o/fferenf  types  of  antennas  and  antenna  transmission  lines  from  the  coaxial  input  line  from  the 
transmitter  or  from  the  antenna  changeover  relay.  Link  L should  be  resonated  with  capacitor 
C at  the  operating  frequency  of  the  transmitter  so  that  the  harmonic  filter  will  operate  into  a 
resistive  load  impedance  of  the  correct  nominal  value. 


ended  output  stage  down  to  the  50-ohni  imped- 
ance of  the  usual  harmonic  filter  and  its  sub- 
sequent load. 

In  a pi  network  of  this  type  the  harmonic 
attenuation  of  the  section  will  be  adequate 
when  the  correct  value  of  Ci  and  L are  being 
used  and  when  the  resonant  dip  in  Cj  is 
sharp.  If  the  dip  in  Ci  is  broad,  or  if  the  plate 
current  persists  in  being  too  high  with  C,  at 
maximum  setting,  it  means  that  a greater  value 


of  capacitance  is  required  at  C,,  assuming  that 
the  values  of  Cj  and  L are  correct. 

22-11  Antenna  Couplers 

As  stated  in  the  previous  section,  an  anten- 
na coupler  is  not  requited  when  the  impedance 
of  the  antenna  transmission  line  is  the  same 
as  the  nominal  impedance  of  the  harmonic  fil- 


HANDBOOK 


Antenna  Couplers  455 


ter,  assuming  that  the  antenna  feed  line  is  be- 
ing operated  with  a low  standing-wave  ratio. 
However,  there  are  many  cases  where  it  is  de- 
sirable to  feed  a multi-band  antenna  from  the 
output  of  the  harmonic  filter,  where  a tuned 
line  is  being  used  to  feed  the  antenna,  or 
where  a long  wire  without  a separate  feed  line 
is  to  be  fed  from  the  output  of  the  harmonic 
filter.  In  such  cases  an  antenna  coupler  is  re- 
quired. 

Some  harmonic  attenuation  will  be  provided 
by  the  antenna  coupler,  particularly  if  it  is 
well  shielded.  In  certain  cases  when  a pi  net- 
work is  being  used  at  the  output  of  the  trans- 
mitter, the  addition  of  a shielded  antenna  cou- 
pler will  provide  sufficient  harmonic  attenua- 
tion. But  in  all  normal  cases  it  will  be  neces- 
sary to  include  a harmonic  filter  between  the 
output  of  the  transmitter  and  the  antenna  cou- 
pler. When  an  adequate  harmonic  filter  is  be- 
ing used,  it  will  not  be  necessary  in  normal 
cases  to  shield  the  antenna  couplet,  except 
from  the  standpoint  of  safety  or  convenience. 

Function  of  on  The  function  of  the  antenna 
Antenna  Coupler  coupler  is,  basically,  to 
transform  the  impedance  of 
the  antenna  system  being  used  to  the  correct 
value  of  resistive  impedance  for  the  harmonic 
filter,  and  hence  lor  the  transmitter.  Thus  the 
antenna  coupler  may  be  used  to  resonate  the 
feeders  or  the  radiating  portion  of  the  antenna 
system,  in  addition  to  its  function  of  imped- 
ance transformation. 

It  is  important  to  remember  that  there  is 
nothing  that  can  be  done  at  the  antenna  cou- 
pler which  will  eliminate  standing  waves  on 
the  antenna  transmission  line.  Standing  waves 
are  the  result  of  reflection  from  the  antenna, 
and  the  coupler  can  do  nothing  about  this  con- 
dition. However,  the  antenna  coupler  can  reso- 
nate the  feed  line  (by  introducing  a conjugate 
impedance)  in  addition  to  providing  an  imped- 
ance transformation.  Thus,  a resistive  imped- 
ance of  the  correct  value  can  be  presented  to 
the  harmonic  filter,  as  in  figure  36,  regardless 
of  any  reasonable  value  of  standing-wave  ratio 
on  the  antenna  transmission  line. 

Types  of  All  usual  types  of  antenna 

Antenna  Couplers  couplers  fall  into  two  clas- 
sifications: (1)  inductively 
coupled  resonant  systems  as  exemplified  by 
those  shown  in  figure  39,  and  (2)  conductively 
coupled  pi-network  systems  such  as  shown  in 
figure  40.  The  inductively-coupled  system  is 
much  more  commonly  used,  since  it  is  conven- 
ient for  feeding  a balanced  line  from  the  co- 
axial output  of  the  usual  harmonic  filter.  The 
pi-network  system  is  most  useful  for  feeding 
a length  of  wire  from  the  output  of  a trans- 
mitter. 


COAX.  TO  SINGLE  WIRE 

RECEIVER  ANTENNA 


ANTENNA  COUPLER 


An  arrangement  such  as  iNustrated 
above  is  convenient  for  feeding  an 
end‘fed  Hertz  antenna^  or  a random 
length  of  wire  for  portable  or  emer- 
gency operationf  from  the  nominal 
value  of  impedance  of  the  harmonic 
filter. 


Several  general  methods  for  using  the  induc- 
tively-coupled resonant  type  of  antenna  cou- 
pler are  illustrated  in  figure  39.  The  coupling 
between  the  link  coil  L and  the  main  tuned  cir- 
cuit need  not  be  variable;  in  fact  it  is  prefer- 
able that  the  correct  link  size  and  placement 
be  determined  for  the  tank  coil  which  will  be 
used  for  each  band,  and  then  that  the  link  be 
made  a portion  of  the  plug-in  coil.  Capacitor 
C then  can  be  adjusted  to  a pre-detetmined 
value  for  each  band  such  that  it  will  resonate 
with  the  link  coil  for  that  band.  The  reactance 
of  the  link  coil  (and  hence  the  reactance  of  the 
capacitor  setting  which  will  resonate  the  coil) 
should  be  about  3 or  4 times  the  impedance  of 
the  transmission  line  between  the  aiitenna  cou- 
pler and  the  harmonic  filter,  so  that  the  link 
coupling  circuit  will  have  an  operating  Q of 
3 or  4.  The  use  of  capacitor  C to  resonate  with 
the  inductance  of  the  link  coil  L will  make  it 
easier  to  provide  a low  standing-wave  ratio 
to  the  output  of  the  harmonic  filter,  simply  by 
adjustment  of  the  antenna-coupler  tank  circuit 
to  resonance.  If  this  capacitor  is  not  included, 
the  system  still  will  operate  satisfactorily,  but 
the  tank  circuit  will  have  to  be  detuned  slight- 
ly from  resonance  so  as  to  cancel  the  induc- 
tive reactance  of  the  coupling  link  and  thus 
provide  a resistive  load  to  the  output  of  the 
harmonic  filter.  Variations  in  the  loading  of 
the  final  amplifier  should  be  made  by  the  cou- 
pling adjustment  at  the  final  amplifier,  not  at 
the  antenna  coupler. 

The  pi-network  type  of  antenna  couplet,  as 
shown  in  figure  40  is  useful  for  certain  appli- 
cations, but  is  primarily  useful  in  feeding  a 
single-wire  antenna  from  a low-impedance 
transmission  line.  In  such  an  application  the 
operating  Q of  the  pi  network  may  be  somewhat 
lower  than  that  of  a pi  network  in  the  plate  cir- 
cuit of  the  final  amplifier  of  a transmitter,  as 
shown  in  figure  38.  An  operating  Q of  3 or  4 


456  Antennas  and  Antenna  Matching 


THE  RADIO 


PARALLEL-WIRE  TO  TO  COAX.  LINES  TO 

«-SO  M,  ANTENNA  RECEIVER  tOM.ANT  20M.AMT. 


Figure  41 

ALTERNATIVE  COAXIAL  ANTENNA 
COUPLER 

This  circuit  is  recommended  not  only  as  be- 
ing most  desirable  when  coaxial  lines  with 
low  s.w-r.  are  being  used  to  feed  antenna 
systems  such  as  rotatable  beams,  but  when 
It  also  is  desired  to  feed  through  open-wire 
line  to  some  sort  of  multi-band  antenna  for 
the  lower  frequency  ranges.  The  tuned  cir- 
cuit of  the  antenna  coupler  is  operative  only 
when  using  the  open-wire  feed,  and  then  It 
is  in  operation  both  for  transmit  and  receive. 


in  such  an  application  will  be  found  to  be  ade- 
quate, since  harmonic  attenuation  has  been 
accomplished  ahead  of  the  antenna  coupler. 
However,  the  circuit  will  be  easier  to  tune, 
although  it  will  not  have  as  great  a bandwidth, 
if  the  operating  Q is  made  higher. 

An  alternative  arrangement  shown  in  figure 
41  utilizes  the  antenna  coupling  tank  circuit 
only  when  feeding  the  coaxial  output  of  the 
transmitter  to  the  open-wire  feed  line  (or  simi- 
lar multi-band  antenna)  of  the  40-  80  meter  an- 
tenna. The  coaxial  lines  to  the  10-meter  beam 
and  to  the  20-meter  beam  would  be  fed  directly 
from  the  output  of  the  coaxial  antenna  change- 
over relay  through  switch  S. 


22-12  A Single-Wire  Antenna  Tuner 

One  of  the  simplest  and  least  expensive 
antennas  for  transmission  and  reception  is 
the  single  wire,  end-fed  Hertz  antenna.  When 
used  over  a wide  range  of  frequencies,  this 
type  of  antenna  exhibits  a very  great  range  of 
input  impedance.  At  the  low  frequency  end  of 
the  spectrum  such  an  antenna  may  present  a 
resistive  load  of  less  than  one  ohm  to  the 
transmitter,  combined  with  a large  positive  or 
negative  value  of  reactance.  As  the  frequency 
of  operation  is  raised,  the  resistive  load  may 


SmQLE  WIRE 
ANTENNA 


Figure  42 

ANTENNA  TUNER  AND  SWR 
INDICATOR  FOR  RANDOM 
LENGTH  HERTZ  ANTENNA 


rise  to  several  thousand  ohms  (near  half-wave 
resonance)  and  the  reactive  component  of  the 
load  can  rapidly  change  from  positive  to  nega- 
tive values,  or  vice-versa. 

It  is  possible  to  match  a 52-ohm  trans- 
mission line  to  such  an  antenna  at  almost  any 
frequency  between  1.8  me  and  30  me  with  the 
use  of  a simple  tuner  of  the  type  shown  in 
figure  42.  A variable  series  inductor  L,  and  a 
variable  shunt  capacitor  Cl  permit  circuit 
resonance  and  impedance  transformation  to 
be  established  for  most  antenna  lengths. 
Switch  SI  permits  the  selection  of  series 
capacitor  C for  those  instances  when  the 
single  wire  antenna  exhibits  large  values  of 
positive  reactance. 

To  provide  indication  for  the  tuning  of  the 
network,  a radio  frequency  bridge  (SWR  meter) 
is  included  to  indicate  the  degree  of  mis- 
match (standing  wave  ratio)  existing  at  the 
input  to  the  tuner.  AH  adjustments  to  the 
tuner  are  made  with  the  purpose  of  reaching 
unity  standing  wave  ratio  on  the  coaxial  feed 
system  between  the  tuner  and  the  transmitter. 

A Procficul  A simple  antenna  tuner  for 

Antenna  Tuner  use  with  transmitters  of 

250  watts  power  or  less 
is  shown  in  figures  43  through  46.  A SWR 
bridge  circuit  is  used  to  indicate  tuner  re- 
sonance. The  resistive  arm  of  the  bridge  con- 
sists of  ten  10-ohm,  1-watt  carbon  resistors 
connected  in  parallel  to  form  a 1-ohm  resistor 
(Rl).  The  other  pair  of  bridge  arms  are  ca- 
pacitive rather  than  resistive.  The  bridge 
detector  is  a simple  r-f  voltmeter  employing  a 
IN56  crystal  diode  and  a 0-1  d.c.  milliammeter. 
A sensitivity  control  is  incorporated  to  prevent 
overloading  the  meter  when  power  is  first 
applied  to  the  tuner.  Final  adjustments  ate 
made  with  the  sensitivity  control  at  its  maxi- 


HANDBOOK 


Single-wire  Antenna 


Tuner  457 


Figure  43 

ANTENNA  TUNER  IS  HOUSED 


Li- 35  TURNS  *18,  2*  DIA., 
3.5" LONG  {AIR-DUX) 
TAP  AT  15  T.,  27  T., 
FROM  POINT  A 


Z\-JOHNSON  35DEZO 
OX-CENTRALAB  TYPE  SEE 
Jl-TYPE  50-239  RECEPTACLE 


IN  METAL  CABINET  7"  x 8"  IN 
SIZE. 


V-E- JOHNSON  ZE9-Z01  VARIABLE 
INDUCTOR  {lOJJH) 


Ri-ten  io-ohm  1-watt  car- 
bon RESISTORS  IN  PARA- 
LLEL. IRC  TYPE  BTA 


Inductance  switch  SI  and  sensi- 
tivity control  are  at  left  with 
counter  dial  for  L2  at  center. 
Output  tuning  capacitor  Cl  is  at 
right,  SVfR  meter  is  mounted 
above  SI, 


Figure  44 

SCHEMATIC,  SINGLE-WIRE 
ANTENNA  TUNER 


mum  (clockwise)  position.  The  bridge  is 
balanced  when  the  input  impedance  of  the 
tuner  is  52  ohms  resistive.  This  is  the  con- 
dition for  maximum  energy  transfer  between 
transmission  line  and  antenna.  The  meter  is 
graduated  in  arbitrary  units,  since  actual  SWR 
value  is  not  required. 


Tuner  Major  parts  placement  in 

Construction  the  tuner  is  shown  in 

figures  43  and  45-  Tapped 
coil  LI  is  mounted  upon  J^-inch  ceramic  in- 
sulators, and  all  major  components  are  mounted 
above  deck  with  the  exception  of  the  SWR 
bridge  (figure  46).  The  components  of  the 
bridge  are  placed  below  deck,  adjacent  to 
the  coaxial  input  plug  mounted  on  the  rear 
apron  of  the  chassis.  The  ten  10-ohm  resistors 
are  soldered  to  two  1-inch  tings  made  of  copper 
wire  as  shown  in  the  photograph.  The  bridge 
capacitors  ate  attached  to  this  assembly  with 
extremely  short  leads. The  1N56  crystal  mounts 
at  right  angles  to  the  resistors  to  insure  mini- 
mum amount  of  capacitive  coupling  between  the 
resistors  and  the  detector.  The  output  lead 
from  the  bridge  passes  through  a ceramic  feed- 
thru  insulator  to  the  top  side  of  the  chassis. 

Connection  to  the  antenna  is  made  by  means 
of  a large  feedthru  insulator  mounted  on  the 
back  of  the  tuner  cabinet.  This  insulator  is 
not  visible  in  the  photographs. 


Bridge  The  SWR  bridge  must  be 

Calibration  Calibrated  for  52  ohm  ser- 

vice. This  can  be  done  by 
temporarily  disconnecting  the  lead  between  the 
bridge  and  the  antenna  tuner  and  connecting  a 
2-watt,  52  ohm  carbon  resistor  to  the  junction 
of  R1  and  the  negative  terminal  of  the  1N56 
diode.  The  opposite  lead  of  the  carbon  resistor 
is  grounded  to  the  chassis  of  the  bridge.  A 
small  amount  of  r-f  energy  is  fed  to  the  input 
of  the  bridge  until  a reading  is  obtained  on 
the  t-f  voltmeter.  The  25  mmfd  bridge  balancing 
capacitor  C2  (see  figure  46)  is  then  adjusted 
with  a fibre-blade  screwdriver  until  a zero 
reading  is  obtained  on  the  meter.  The  sensi- 
tivity control  is  advanced  as  the  meter  null 
grows,  in  order  to  obtain  the  exact  point  of 
bridge  balance.  When  this  point  is  found,  the 
carbon  resistor  should  be  removed  and  the 
bridge  attached  to  the  antenna  tuner.  The 
bridge  capacitor  is  sealed  with  a drop  of  nail 
polish  to  prevent  misadjustment. 

Tuner  All  tuning  adjustments  ate 

Adjustments  made  to  obtain  proper 

transmitter  loading  with  a 
balanced  (zero  meter  reading)  bridge  condition. 
The  tuner  is  connected  to  the  transmitter 
through  a random  length  of  52  ohm  coaxial 
line,  and  the  single  wire  antenna  is  attached 
to  the  output  terminal  of  the  tuner.  Transmitter 
loading  controls  are  set  to  approximate  a 52 


F igure  45 


SltlUlllf 


REAR  VIEW  OF 
TUNER  SHOWING 
PLACEMENT  OF 
MAJOR  COMPON- 
ENTS 

Rotary  inductor  is  driv- 
en  by  Johnson  116-208’ 
4 counter  dial.  Coaxial 
input  receptacle  J1 
Is  mounted  directly  be- 
low rotary  inductor. 


ohm  termination.  The  transmitter  is  turned  on 
(preferably  at  reduced  input)  and  resonance  is 
established  in  the  amplifier  tank  circuit.  The 
sensitivity  control  of  the  tuner  is  adjusted  to 
provide  near  full  scale  deflection  on  the  bridge 
meter.  Various  settings  of  SI,  L2,  and  Cl 
should  be  tried  to  obtain  a reduction  of  bridge 
reading.  As  tuner  resonance  is  approached, 
the  meter  reading  will  decrease  and  the  sensi- 
tivity control  should  be  advanced.  When  the 
system  is  in  resonance,  the  meter  will  read 
zero.  All  loading  adjustments  may  then  be 
made  with  the  transmitter  controls.  The  tuner 
should  be  readjusted  whenever  the  frequency 
of  the  transmitter  is  varied  by  an  appreciable 
amount. 


Figure  46 

CLOSE-UP  OF  SWR  BRIDGE 
Simple  SWR  bridge  is  mounted 
below  the  chassis  of  the  tuner. 
Carbon  resistors  are  mounted  to 
fwo  copper  rings  to  form  low 
inductance  one-ohm  resistor. 
Bridge  capacitors  form  triangular 
configuration  for  lowest  lead 
inductance.  Balancing  capacitor 
C2  is  at  lower  right. 


CHAPTER  TWENTY-THREE 


High  Frequency  Antenna  Arrays 


It  is  becoming  of  increasing  importance  in 
most  types  of  radio  communication  to  be  cap- 
able of  concentrating  the  radiated  signal  from 
the  transmitter  in  a certain  desired  direction 
and  to  be  able  to  discriminate  at  the  receiver 
against  reception  from  directions  other  than 
the  desired  one.  Such  capabilities  involve  the 
use  of  directive  antenna  arrays. 

Few  simple  anrennas,  except  the  single  ver- 
tical element,  radiate  energy  equally  well  in 
all  azimuth  (horizontal  or  compass)  direcrions. 
All  horizontal  antennas,  except  those  specifi- 
cally designed  to  give  an  omnidirectional  azi- 
muth radiation  pattern  such  as  the  turnstile, 
have  some  directive  properties.  These  proper- 
ties depend  upon  the  length  of  the  antenna  in 
wavelengths,  the  height  above  ground,  and  the 
slope  of  the  radiator. 

The  various  forms  of  the  half-wave  horizon- 
tal antenna  produce  maximum  radiation  at  tight 
angles  to  the  wire,  but  the  directional  effect 
is  not  great.  Nearby  objects  also  minimize  the 
directivity  of  a dipole  radiator,  so  that  it  hard- 
ly seems  worth  while  to  go  to  the  trouble  to 
rotate  a simple  half-wave  dipole  in  an  attempt 
to  improve  transmission  and  reception  in  any 
direction. 

The  half-wave  doublet,  folded  dipole,  zepp, 
single-wire-fed,  matched  impedance,  and  John- 
son Q antennas  all  have  practically  the  same 
radiation  pattern  when  properly  built  and  ad- 


justed. They  all  are  dipoles,  and  the  feeder 
system,  if  it  does  not  radiate  in  itself,  will 
have  no  effect  on  the  radiation  pattern. 


23-1  Directive  Antennas 

When  a multipliciry  of  radiating  elements  is 
located  and  phased  so  as  to  reinforce  the  radi- 
ation in  certain  desired  directions  and  to  neu- 
tralize radiation  in  other  directions,  a direc- 
tive antenna  array  is  formed. 

The  function  of  a directive  antenna  when 
used  for  transmitting  is  to  give  an  increase  in 
signal  strength  in  some  direction  at  the  ex- 
pense of  radiation  in  other  directions.  For  re- 
ception, one  might  find  useful  an  antenna  giv- 
ing little  or  no  gain  in  the  direction  from  which 
it  is  desired  to  receive  signals  if  the  antenna 
is  able  to  discriminate  against  interfering  sig- 
nals and  static  arriving  from  orher  directions. 
A good  directive  transmitting  antenna,  however, 
can  also  be  used  to  good  advantage  for  recep- 
tion. 

If  radiation  can  be  confined  to  a narrow 
beam,  the  signal  intensity  can  be  increased  a 
great  many  times  in  the  desired  direction  of 
transmission.  This  is  equivalent  to  increasing 
the  power  output  of  the  transmitter.  On  the 
higher  frequencies,  it  is  more  economical  to 


459 


46  0 High  Frequency  Directive  Antennas 


THE  RADIO 


O 10  30  SO  too  300  SCO  tOOO  3000  fOOOO 

GREAT  CIRCLE  DISTANCE  IN  MILES 

Figure  1 

OPTIMUM  ANGLE  OF  RADIATION 
WITH  RESPECT  TO  DISTANCES 

Shown  above  is  a plot  of  the  optimum  angle 
of  rodiation  for  one-hop  and  two-hop  com- 
munication. An  operating  frequency  close  to 
the  optimum  working  frequency  for  the  com- 
municotion  distance  is  assumed. 


use  a directive  antenna  than  to  increase  trans- 
mitter power,  if  more  than  a few  watts  of  power 
is  being  used. 

Directive  antennas  for  the  high-frequency 
range  have  been  designed  and  used  commer- 
cially with  gains  as  high  as  23  db  over  a sim- 
ple dipole  radiator.  Gains  as  high  as  35  db  are 
common  in  direct-ray  microwave  communication 
and  radar  systems.  A gain  of  23  db  represents 
a power  gain  of  200  times  and  a gain  of  35  db 
represents  a power  gain  of  almost  3500  times. 
However,  an  antenna  with  a gain  of  only  15  to 
20  db  is  so  sharp  in  its  radiation  pattern  that 
it  is  usable  to  full  advantage  only  for  point- 
to-point  work. 

The  increase  in  radiated  power  in  the  de- 
sired direction  is  obtained  at  the  expense  of 
radiation  in  the  undesired  directions.  Power 
gains  of  3 to  12  db  seem  to  be  most  practi- 
cable for  amateur  communication,  since  the 
width  of  a beam  with  this  order  of  power  gain 
is  wide  enough  to  sweep  a fairly  large  area. 
Gains  of  3 to  12  db  represent  effective  trans- 
mitter power  increases  from  2 to  16  times. 

Horizontal  Pattern  There  is  a certain  optimum 
vs.  Vertical  Angle  vertical  angle  of  radiation 
for  sky-wave  communica- 
tion, this  angle  being  dependent  upon  distance, 
frequency,  time  of  day,  etc.  Energy  radiated  at 
an  angle  much  lower  than  this  optimum  angle 
is  largely  lost,  while  radiation  at  angles  much 


Figure  2 

FREE-SPACE  FIELD  PATTERNS  OF 
LONG-WIRE  ANTENNAS 

The  presence  of  the  earth  distorts  the  field 
pattern  in  such  a manner  that  the  azimuth 
pattern  becomes  a function  of  the  elevation 
angle. 


higher  than  this  optimum  angle  oftentimes  is 
not  nearly  so  effective. 

For  this  reason,  the  horizontal  directivity 
pattern  as  measured  on  the  ground  is  of  no  im- 
port when  dealing  with  frequencies  and  dis- 
tances dependent  upon  sky-wave  propagation. 
It  is  the  horizontal  directivity  (or  gain  or  dis- 
crimination) measured  at  the  most  useful  ver- 
tical angles  of  radiation  that  is  of  conse- 
quence. The  horizontal  radiation  pattern;  as 
measured  on  the  ground,  is  considerably  differ- 
ent from  the  pattern  obtained  at  a vertical 
angle  of  15°,  and  still  more  different  from  a 
pattern  obtained  at  a vertical  angle  of  30°. 
In  general,  the  energy  which  is  radiated  at 
angles  higher  than  approximately  30°  above 
the  earth  is  effective  at  any  frequency  only  for 
local  work. 

For  operation  at  frequencies  in  the  vicinity 
of  14  Me.,  the  most  effective  angle  of  radiation 
is  usually  about  15°  above  the  horizon,  from 
any  kind  of  antenna.  The  most  effective  angles 
for  10-meter  operation  are  those  in  the  vicinity 
of  10°.  Figure  1 is  a chart  giving  the  optimum 
vertical  angle  of  radiation  for  sky-wave  propa- 
gation in  terms  of  the  great-circle  distance  be- 
tween the  transmitting  and  receiving  antennas. 


HANDBOOK 


Long  Wire  Radiators  461 


Figure  3 

DIRECTIVE  GAIN  OF 
LONG-WIRE  ANTENNAS 


Types  of  There  is  an  enormous  vari- 

Directlve  Arrays  ety  of  directive  antenna  ar- 
rays that  can  give  a substan- 
tial power  gain  inthedesited  direction  of  trans- 
mission or  reception.  However,  some  are  more 
effective  than  others  requiting  the  same  space. 
In  general  it  may  be  stated  that  long-wire  an- 
tennas of  various  types,  such  as  the  single 
long  wire,  the  V beam,  and  the  rhombic,  ate 
less  effective  for  a given  space  than  arrays 
composed  of  resonant  elements,  but  the  long- 
wire  arrays  have  the  significant  advantage 
that  they  may  be  used  over  a relatively  large 
frequency  range  while  resonant  arrays  are  usa- 
ble only  over  a quite  narrow  frequency  band. 


23-2  Long  Wire  Radiators 

Harmonically  operated  long  wires  radiate 
better  in  certain  directions  than  others,  but 
cannot  be  considered  as  having  appreciable 
directivity  unless  several  wavelengths  long. 
The  current  in  adjoining  half-wave  elements 
flows  in  opposite  directions  at  any  instant, 
and  thus,  the  radiation  from  the  various  ele- 
ments adds  in  certain  directions  and  cancels 
in  others. 

A half-wave  doublet  in  free  space  has  a 
"doughnut”  of  radiation  surrounding  it.  A full 
wave  has  2 lobes,  3 half  waves  3,  etc.  When 
the  radiator  is  made  more  than  4 half  wave- 
lengths long,  the  end  lobes  (cones  of  radia- 
tion) begin  to  show  noticeable  power  gain  over 
a half-wave  doublet,  while  the  broadside  lobes 
get  smaller  and  smaller  in  amplitude,  even 
though  numerous  (figure  2). 


The  horizontal  radiation  pattern  of  such  an- 
tennas depends  upon  the  vertical  angle  of  radi- 
ation being  considered.  If  the  wire  is  more 
than  4 wavelengths  long,  the  maximum  radia- 
tion at  vertical  angles  of  15“  to  20°  (useful 
for  dx)  is  in  line  with  the  wire,  being  slightly 
greater  a few  degrees  either  side  of  the  wire 
than  directly  off  the  ends.  The  directivity  of 
the  main  lobes  of  radiation  is  not  particularly 
sharp,  and  the  minor  lobes  fill  in  between  the 
main  lobes  to  permit  working  stations  in  neat- 
ly all  directions,  though  the  power  radiated 
broadside  to  the  radiator  will  not  be  great  if 
the  radiator  is  more  than  a few  wavelengths 
long.  The  directive  gain  of  long-wire  antennas, 
in  terms  of  the  wire  length  in  wavelengths  is 
given  in  figure  3. 

To  maintain  the  out-of-phase  condition  in 
adjoining  half-wave  elements  throughout  the 
length  of  the  radiator,  it  is  necessary  that  a 
harmonic  antenna  be  fed  either  at  one  end  or 
at  a current  loop.  If  fed  at  a voltage  loop,  the 
adjacent  sections  will  be  fed  in  phase,  and  a 
different  radiation  pattern  will  result. 

The  directivity  of  a long  wire  does  not  in- 
crease very  much  as  the  length  is  increased 
beyond  about  15  wavelengths.  This  is  due  to 
the  fact  that  all  long-wire  antennas  are  ad- 
versely affected  by  the  r-f  resistance  of  the 
wire,  and  because  the  current  amplitude  be- 
gins to  become  unequal  at  different  current 
loops,  as  a result  of  attenuation  along  the  wire 
caused  by  radiation  and  losses.  As  the  length 
is  increased,  the  tuning  of  the  antenna  be- 
comes quite  broad.  In  fact,  a long  wire  about 
15  waves  long  is  practically  aperiodic,  and 
works  almost  equally  well  over  a wide  range 
of  frequencies. 


462  High  Frequency  Directive  Antennas 


THE  RADIO 


LONG- ANTENNA  DESIGN  TABLE 

APPROXIMATE  LENGTH  IN  FEET END-FED  ANTENNAS 

FREQUENCY 

IN  MC. 

1A 

2X 

3X 

4X 

4ix 

29 

33 

50 

67 

64 

101 

116 

135 

1 52 

26 

34 

52 

69 

67 

104 

122 

140 

157 

21.4 

45 

66 

91  t/2 

114  1/2 

136  1/2 

160  1/2 

165  1/2 

209  1/2 

21.2 

45  V4 

66  1/4 

91  3/4 

114  3/4 

136  3t4 

160  3/4 

165  3M 

209  3/4 

21.0 

45  1/2 

66  1/2 

92 

115 

137 

161 

166 

210 

14.2 

6 7 1/2 

102 

1 37 

171 

206 

240 

275 

310 

14.  0 

66  1/2 

103  1/2 

139 

174 

209 

244 

279 

314 

7.3 

136 

206 

276 

346 

416 

466 

555 

625 

7, 15 

1 36  1/2 

207 

277 

34  7 

417 

487 

557 

627 

7.0 

137 

207  1/2 

277  t/2 

346 

416 

466 

558 

626 

4.0 

240 

362 

485 

618 

730 

653 

977 

1100 

3.6 

252 

361 

51  1 

640 

770 

900 

1030 

1 160 

3.6 

266 

403 

540 

676 

612 

950 

1090 

1220 

3.5 

274 

414 

555 

696 

635 

977 

1 1 20 

2.0 

460 

725 

972 

1230 

1475 

1.9 

504 

763 

1020 

1260 

1.6 

532 

805 

1060 

One  of  the  most  practical  methods  of  feed- 
ing a long-wire  antenna  is  to  bring  one  end  of 
it  into  the  radio  room  for  direct  connection  to 
a tuned  antenna  circuit  which  is  link-coupled 
through  a harmonic-attenuating  filter  to  the 
transmitter.  The  antenna  can  be  tuned  effec- 
tively to  resonance  for  operation  on  any  har- 
monic by  means  of  the  tuned  circuit  which  is 
connected  to  the  end  of  the  antenna.  A ground 
is  sometimes  connected  to  the  center  of  the 
tuned  coil. 

If  desired,  the  antenna  can  be  opened  and 
current-fed  at  a point  of  maximum  current 
means  of  low-impedance  ribbon  line,  or  by  a 
quarter-wave  matching  section  and  open  line. 

23-3  The  V Antenna 

If  two  long-wire  antennas  are  built  in  the 


form  of  a V,  it  is  possible  to  make  two  of  the 
maximum  lobes  of  one  leg  shoot  in  the  same 
direction  as  two  of  the  maximum  lobes  of  the 
other  leg  of  the  V.  The  resulting  antenna  is 
bidirectional  (two  opposite  directions)  for  the 
main  lobes  of  radiation.  Each  side  of  the  V 
can  be  made  any  odd  or  even  number  of  quarter 
wavelengths,  depending  on  the  method  of  feed- 
ing the  apex  of  the  V.  The  complete  system 
must  be  a multiple  of  half  waves.  If  each  leg 
is  an  even  number  of  quarter  waves  long,  the 
antenna  must  be  voltage-fed  at  the  apex;  if  an 
odd  number  of  quarter  waves  long,  current  feed 
must  be  used. 

By  choosing  the  proper  apex  angle,  figure 
4 and  figure  5,  the  lobes  of  radiation  from  the 
two  long-wire  antennas  aid  each  other  to  form 
a bidirectional  beam.  Each  wire  by  itself 
would  have  a radiation  pattern  similar  to  that 


Figure  4 

INCLUDED  ANGLE  FOR  A 
"V"  BEAM 

Showing  the  included  angle  be- 
tween the  legs  of  a V beam  for 
various  leg  lengths.  For  opth 
mum  alignment  of  the  radiation 
lobe  at  the  correct  vertical 
angle  with  leg  lengths  less  than 
three  wavelengths^  the  optimum 
Included  angle  Is  shown  by  the 
dashed  curve. 


HANDBOOK 


The  V Antenna  463 


Figure  5 

TYPICAL  "V"  BEAM  ANTENNA 


for  a long  wire.  The  reaction  of  one  upon  the 
other  removes  two  of  the  four  main  lobes,  and 
increases  the  other  two  in  such  a way  as  to 
form  two  lobes  of  still  greater  magnitude. 

The  correct  wire  lengths  and  the  degree  of 
the  angle  S are  listed  in  the  V- Antenna  Design 
Table  for  various  frequencies  in  the  10-,  20- 
and  40-meter  amateur  bands.  Apex  angles  for 
all  side  lengths  are  given  in  figure  4.  The 
gain  of  a''V”  beam  in  terms  of  the  side  length 
when  optimum  apex  angle  is  used  is  given  in 
figure  6. 

The  legs  of  a very  long  V antenna  are  usual- 
ly so  arranged  that  the  included  angle  is  twice 
the  angle  of  the  major  lobe  from  a single  wire 
if  used  alone.  This  arrangement  concentrates 
the  radiation  of  each  wire  along  the  bisector 
of  the  angle,  and  permits  part  of  the  other 
lobes  to  cancel  each  other. 

With  legs  shorter  than  3 wavelengths,  the 
best  directivity  and  gain  ate  obtained  with  a 
somewhat  smaller  angle  than  that  determined 
by  the  lobes.  Optimum  directivity  for  a one- 
wave  V is  obtained  when  the  angle  is  90° 


Figure  6 

DIRECTIVE  GAIN  OF  A "V"  BEAM 

This  curve  shows  the  approximate  directive 
gain  of  a V beam  with  respect  to  a half-wave 
antenna  located  the  same  distance  above 
ground,  in  terms  of  the  side  length  L. 


rather  than  180°,  as  determined  by  the  ground 
pattern  alone. 

If  very  long  wires  are  used  in  the  V,  the 
angle  between  the  wires  is  almost  unchanged 
when  the  length  of  the  wires  in  wavelengths 
is  altered.  However,  an  error  of  a few  degrees 
causes  a much  larger  loss  in  directivity  arid 
gain  in  the  case  of  the  longer  V than  in  the 
shorter  one. 

The  vertical  angle  at  which  the  wave  is 
best  transmitted  or  received  from  a horizontal 
V antenna  depends  largely  upon  the  included 
angle.  The  sides  of  the  V antenna  should  be 
at  least  a half  wavelength  above  ground;  com- 
mercial practice  dictates  a height  of  approxi- 
mately a full  wavelength  above  ground. 


V-ANTENNA  DESIGN  TABLE 

FREQUENCY 

L =X 

L = 2X 

L =4X 

L =8X 

1 N K 1 LOCYCLES 

=90* 

6 ~ 70* 

6 - 52* 

39* 

28000 

ZA'B" 

69'B'' 

140' 

260' 

29000 

33'e* 

67 '3" 

135' 

271  ' 

21100 

45'9" 

91'9" 

183' 

366' 

21300 

45*4" 

91'4  - 

182'  6' 

365' 

14050 

89' 

139' 

279' 

558' 

14150 

68'6" 

136' 

277' 

555' 

14250 

68'  2' 

137' 

275' 

552' 

7020 

138'2'' 

278' 

556' 

1120' 

7100 

136'8* 

275' 

552' 

1106' 

7200 

1 34' 10" 

271 ' 

545' 

1090' 

464  High  Frequency  Directive  Antennas 


THE  RADIO 


23-4  The  Rhombic 

Antenna 

The  terminated  rhombic  or  diamond  is  prob- 
ably the  most  effective  directional  antenna 
that  is  practical  for  amateur  communication. 
This  antenna  is  non-resonant,  with  the  result 
that  it  can  be  used  on  three  amateur  bands, 
such  as  10,  20,  and  40  meters.  When  the  an- 
tenna is  non-resonant,  i.e.,  properly  termi- 
nated, the  system  is  undirectional,  and  the 
wire  dimensions  are  not  critical. 

Rhombic  When  the  free  end  is  terminated 

Termination  with  a resistance  of  a value 
between  700  and  800  ohms  the 
backwave  is  eliminated,  the  forward  gain  is 
increased,  and  the  antenna  can  be  used  on 
several  bands  without  changes.  The  terminat- 
ing resistance  should  be  capable  of  dissipat- 
ing one-third  the  power  output  of  the  trans- 
mitter, and  should  have  very  little  reactance. 
For  medium  or  low  power  transmitters,  the 
non-inductive  plaque  resistors  will  serve  as 
a satisfactory  termination.  Several  manufactur- 
ers offer  special  resistors  suitable  for  termi- 
nating a rhombic  antenna.  The  terminating  de- 
vice should,  for  technical  reasons,  present  a 
small  amount  of  inductive  reactance  at  the 
point  of  termination. 

A compromise  terminating  device  commonly 
used  consists  of  a terminated  250-foot  or 
longer  length  of  line,  made  of  resistance  wire 
which  does  not  have  too  much  resistance  per 
unit  length.  If  the  latter  qualification  is  not 
met,  the  reactance  of  the  line  will  be  exces- 
sive. A 250-foot  line  consisting  of  no.  25 
nichrome  wire,  spaced  6 inches  and  terminated 
with  800  ohms,  will  serve  satisfactorily.  Be- 
cause of  the  attenuation  of  the  line,  the 
lumped  resistance  at  the  end  of  the  line  need 
dissipate  but  a few  watts  even  when  high  pow- 
er is  used.  A half-dozen  5000-ohm  2-watt  car- 
bon resistors  in  parallel  will  serve  for  all  ex- 
cept very  high  power.  The  attenuating  line 
may  be  folded  back  on  itself  to  take  up  less 
room. 

The  determination  of  the  best  value  of  termi- 
nating resistor  may  be  made  while  receiving, 
if  the  input  impedance  of  the  receiver  is  ap- 
proximately 800  ohms.  The  value  of  resistor 
which  gives  the  best  directivity  on  reception 
will  not  give  the  most  gain  when  transmitting, 
but  there  will  be  little  difference  between  the 
two  conditions. 

The  input  resistance  of  the  rhombic  which 
is  reflected  into  the  transmission  line  that 
feeds  it  is  always  somewhat  less  than  the 
terminating  resistance,  and  is  around  700  to 
750  ohms  when  the  terminating  resistor  is  800 
ohms. 


Design  data  is  given  In  terms  of  the  wave 
angle  (vertical  angle  of  transmission  and  re- 
ception) of  the  antenna.  The  lengths  I ore 
for  the  maximum  output”  design;  the  shorter 
lengths  I * are  for  the  ” alignment”  method 
which  gives  approximately  I.S  db  less  gain 
with  a considerable  reduction  In  the  space 
required  for  the  antenna.  The  values  of  side 
length,  tilt  angle,  and  height  for  a given 
wave  angle  are  obtained  by  drawing  a ver- 
tical line  upward  from  the  desired  wave 
angle. 


The  antenna  should  be  fed  with  a non-reso- 
nant line  having  a characteristic  impedance 
of  650  to  700  ohms.  The  four  corners  of  the 
rhombic  should  be  at  least  one-half  wave- 
length above  ground  for  the  lowest  frequency 
of  operation.  For  three-band  operation  the 
proper  tilt  angle  0 for  the  center  band  should 
be  observed. 

The  rhombic  antenna  transmits  a horizon- 
tally-polarized wave  at  a relatively  low  angle 
above  the  horizon.  The  angle  of  radiation 
(wave  angle)  decreases  as  the  height  above 
ground  is  increased  in  the  same  manner  as 
with  a dipole  antenna.  The  rhombic  should 
not  be  tilted  in  any  plane.  In  other  words,  the 
poles  should  all  be  of  the  same  height  and 
the  plane  of  the  antenna  should  be  parallel 
with  the  ground. 


HANDBOOK 


The  Rhombic  Antenna  465 


Figure  8 

TYPICAL  RHOMBIC 
ANTENNA  DESIGN 

The  antenna  system  illus~ 
trated  above  may  be  used 
over  the  frequency  range  from 
7 to  29  Me.  without  change. 
The  directivity  of  the  system 
may  be  reversed  by  the  sys- 
tem discussed  in  the  text. 


A considerable  amount  of  directivity  is  lost 
when  the  terminating  resistor  is  left  off  the 
end  and  the  system  is  operated  as  a resonant 
antenna.  If  it  is  desired  to  reverse  the  direc- 
tion of  the  antenna  it  is  much  better  practice 
to  run  transmission  lines  to  both  ends  of  the 
antenna,  and  then  run  the  terminating  line  to 
the  operating  position.  Then  with  the  aid  of 
two  d-p-d-t  switches  it  will  be  possible  to  con- 
nect either  feeder  to  the  antenna  changeover 
switch  and  the  other  feeder  to  the  terminating 
line,  thus  reversing  the  direction  of  the  array 
and  maintaining  the  same  termination  for 
either  direction  of  operation. 

Figure  7 gives  curves  for  optimum-design 
rhombic  antennas  by  both  the  maximum-out- 
put method  and  the  alignment  method.  The 
alignment  method  is  about  1.5  db  down  from 
the  maximum  output  method  but  requires  only 
about  0.74  as  much  leg  length.  The  height  and 
tilt  angle  is  the  same  in  either  case.  Figure 
8 gives  construction  data  for  a recommended 
rhombic  antenna  for  the  7.0  through  29.7  Me. 


bands.  This  antenna  will  give  about  11  db 
gain  in  the  14.0-Mc.  band.  The  approximate 
gain  of  a rhombic  antenna  over  a dipole,  both 
above  normal  soil,  is  given  in  figure  9- 


23-5  Stacked-Dipole 

Arrays 

The  characteristics  of  a half-wave  dipole 
already  have  been  described.  When  another 
dipole  is  placed  in  the  vicinity  and  excited 
either  directly  or  parasitically,  the  resultant 
radiation  pattern  will  depend  upon  the  spacing 
and  phase  differential,  as  well  as  the  relative 
magnitude  of  the  currents.  With  spacings  less 
than  0.65  wavelength,  the  radiation  is  mainly 
broadside  to  the  two  wires  (bidirectional)  when 
there  is  no  phase  difference,  and  through  the 
wires  (end  fire)  when  the  wires  are  180°  out 
of  phase.  With  phase  differences  between  0° 


Figure  9 

RHOMBIC  ANTENNA  GAIN 

Showing  the  theoretical  gain  of  a rhombic 
antenna,  in  terms  of  the  side  length,  over  a 
half-wave  antenna  mounted  at  the  same 
height  above  the  same  type  of  soil. 


466  High  Frequency  Directive  Antennas 


THE  RADIO 


t^S-H 

■f 

/'  PLANE  OF  WIRES 

I . END  VIEW 


Figure  10 

RADIATION  PATTERNS  OF  A PAIR  OF 
DIPOLES  OPERATING  WITH  IN-PHASE 
EXCITATION,  AND  WITH  EXCITATION 
180®  OUT  OF  PHASE 

If  the  dipoles  are  oriented  horizontally  most 
of  the  directivity  will  be  In  the  vertical 
plane;  if  they  are  oriented  vertically  most 
of  the  directivity  will  be  in  the  horizontal 
plane. 


and  180°  (45°,  90°,  and  135°  for  instance), 
the  pattern  is  unsymmetrical,  the  radiation  be- 
ing greater  in  one  direction  than  in  the  oppo- 
site direction. 

With  spacings  of  more  than  0.8  wavelength, 
more  than  two  main  lobes  appear  for  all  phas- 
ing combinations;  hence,  such  spacings  are 
seldom  used. 

In-Phase  With  the  dipoles  driven  so  as  to 
Spacing  be  in  phase,  the  most  effective 
spacing  is  between  0.5  and  0.7 
wavelength.  The  latter  provides  greater  gain, 
but  minor  lobes  are  present  which  do  not  ap- 
pear at  0.5-wavelength  spacing.  The  radiation 
is  broadside  to  the  plane  of  the  wires,  and  the 
gain  is  slightly  greater  than  can  be  obtained 
from  two  dipoles  out  of  phase.  The  gain  falls 
off  rapidly  for  spacings  less  than  0.375  wave- 
length, and  there  is  little  point  in  using  spac- 
ing of  0.25  wavelength  or  less  with  in-phase 
dipoles,  except  where  it  is  desirable  to  in- 
crease the  radiation  resistance.  (See  Multi- 
Wire  Doublet.) 

Out  of  Phase  When  the  dipoles  are  fed  180° 
Spacing  out  of  phase,  the  directivity  is 

through  the  plane  of  the  wires, 
and  is  greatest  with  close  spacing,  though 
there  is  but  little  difference  in  the  pattern 
after  the  spacing  is  made  less  than  0.125 
wavelength.  The  radiation  resistance  becomes 
so  low  for  spacings  of  less  than  0.1  wave- 
length that  such  spacings  are  not  practicable. 


Li  Le  Lz  Li 


Figure  11 

THE  FRANKLIN  OR  COLINEAR 
ANTENNA  ARRAY 

An  antenna  of  this  type,  regardless  of  the 
number  of  elements,  attains  all  of  its  direc- 
tivity through  sharpening  of  the  horizontal 
or  azimuth  radiation  pattern;  no  vertical  di- 
rectivity is  provided.  Hence  a long  anfenno 
of  this  type  has  an  extremely  sharp  azimuth 
pattern,  but  no  vertical  directivity. 


In  the  three  foregoing  examples,  most  of  the 
directivity  provided  is  in  a plane  at  a tight 
angle  to  the  wires,  though  when  out  of  phase, 
the  directivity  is  in  a line  through  the  wires, 
and  when  in  phase,  the  directivity  is  broadside 
to  them.  Thus,  if  the  wires  are  oriented  verti- 
cally, mostly  horizontal  directivity  will  be 
provided.  If  the  wires  are  oriented  horizontally, 
most  of  the  directivity  obtained  will  be  verti- 
cal directivity. 

To  increase  the  sharpness  of  the  directivity 
in  all  planes  that  include  one  of  the  wires, 
additional  identical  elements  are  added  in  the 
line  of  the  wires,  and  fed  so  as  to  be  in  phase. 
The  familiar  H array  is  one  array  utilizing 
both  types  of  directivity  in  the  manner  pre- 
scribed. The  two-section  Kraus  flat-top  beam 
is  another. 

These  two  antennas  in  their  various  forms 
are  directional  in  a horizontal  plane,  in  addi- 
tion to  being  low-angle  radiators,  and  are  per- 
haps the  most  practicable  of  the  bidirectional 
stacked-dipole  arrays  for  amateur  use.  More 
phased  elements  can  be  used  to  provide  great- 
er directivity  in  planes  including  one  of  the 
radiating  elements.  The  H then  becomes  a 
Sterba- curtain  array. 

For  unidirectional  work  the  most  practicable 
stacked-dipole  arrays  for  amateur-band  use 
are  parasitically-excited  systems  using  rela- 
tively close  spacing  between  the  reflectors 
and  the  directors.  Antennas  of  this  type  are 
described  in  detail  in  a later  chapter.  The 
next  most  practicable  unidirectional  array  is 
an  H or  a Sterba  curtain  with  a similar  system 
placed  approximately  one-quarter  wave  behind. 
The  use  of  a reflector  system  in  conjunction 
with  any  type  of  stacked-dipole  broadside  ar- 
ray will  increase  the  gain  by  3 db. 


HANDBOOK 


Coline  ar  Arrays  467 


COLINEAR  ANTENNA  DESIGN  CHART 

FREQUENCY 

IN  MC. 

Li 

L2 

L3 

26.5 

le'B" 

17' 

S'e* 

21.2 

22'6" 

23'3* 

11'6* 

14.  2 

33'8* 

34'?" 

17'3* 

7.  15 

67' 

ed'd" 

34'4'' 

4,0 

1 20' 

123' 

ai'd" 

3.6 

1 33' 

1 se's* 

66'2‘' 

Colineqr  The  simple  colinear  antenna  array 
Arrays  is  a very  effective  radiating  system 
for  the  3.5-Mc.  and  7.0-Mc.  bands, 
but  its  use  is  not  recommended  on  higher  fre- 
quencies since  such  arrays  do  not  possess 
any  vertical  directivity.  The  elevation  radia- 
tion pattern  for  such  an  array  is  essentially 
the  same  as  for  a half-wave  dipole.  This  con- 
sideration applies  whethet  the  elements  are  of 
normal  length  ot  are  extended. 

The  colinear  antenna  consists  of  two  or 
more  radiating  sections  from  0.5  to  0.65  wave- 
lengths long,  with  the  current  in  phase  in  each 
section.  The  necessary  phase  reversal  between 
sections  is  obtained  through  the  use  of  reso- 
nant tuning  stubs  as  illustrated  in  figure  11. 
The  gain  of  a colinear  array  using  half-wave 
elements  (in  decibels)  is  approximately  equal 
to  the  number  of  elements  in  the  array.  The 
exact  figures  are  as  follows: 

Number  of  Elements  2 3 4 5 6 

Gain  in  Decibels  1.8  3.3  4.5  5.3  6.2 

As  additional  in-phase  colinear  elements 
ate  added  to  a doublet,  the  tadiation  resistance 
goes  up  much  fastet  than  when  additional  half 
waves  are  added  out  of  phase  (harmonic  oper- 
ated antenna). 

For  a colinear  array  of  from  2 to  6 elements. 


— + — 4-  — 4—  (^,-H 


Figure  13 

TWO  COLINEAR  HALF-WAVE  ANTENNAS 
IN  PHASE  PRODUCE  A 3 DB  GAIN  WHEN 
SEPARATED  ONE-HALF  WAVELENGTH 


h ^ H h — — H 

A B 

A- B-I50n  FEED  POINT  GAIN  APPROX.  3DB 

Figure  12 

DOUBLE  EXTENDED  ZEPP  ANTENNA 

For  best  results,  antenna  should  be  tuned  to 
operating  frequency  by  means  of  grid^dip 
oscillator. 


the  terminal  radiation  resistance  in  ohms  at 
any  current  loop  is  approximately  100  times 
the  number  of  elements. 

It  should  be  borne  in  mind  that  the  gain  from 
a colinear  antenna  depends  upon  the  sharpness 
of  the  horizontal  directivity  since  no  vertical 
directivity  is  provided.  An  array  with  several 
colinear  elements  will  give  considerable  gain, 
but  will  have  a sharp  horizontal  radiation 
pattern. 

Double  Extended  The  gain  of  a conventional 
Zepp  two-element  Franklin  colin- 

ear antenna  can  be  increased 
to  a value  approaching  that  obtained  from  a 
three-element  Franklin,  simply  by  making  the 
two  radiating  elements  230°  long  instead  of 
180°  long.  The  phasing  stub  is  shortened  cor- 
respondingly to  maintain  the  whole  array  in 
resonance.  Thus,  instead  of  having  0.5-wave- 
length elements  and  0.25-wavelength  stub,  the 
elements  are  made  0.64  wavelength  long  and 
the  stub  is  approximately  0.11  wavelength 
long. 

Dimensions  for  the  double  extended  Zepp 
are  given  in  figure  12. 

The  vertical  directivity  of  a colineat  anten- 
na having  230°  elements  is  the  same  as  for 
one  having  180°  elements.  There  is  little  ad- 
vantage in  using  extended  sections  when  the 
total  length  of  the  array  is  to  be  greater  than 
about  1.5  wavelength  overall  since  the  gain 


Figure  14 

PRE-CUT  LINEAR  ARRAY  FOR  40.METER 
OPERATION 


46  8 High  Frequency  Directive  Antennas 


THE  RADIO 


of  a colinear  antenna  is  proportional  to  the 

overall  length,  whether  the  individual  radiating 
elements  are  ]4  wave,  wave  or  Va  wave  in 
length. 

Spaced  Half  The  gain  of  two  colinear  half 

Wave  Antennas  waves  may  be  increased  by 

increasing  the  physical  spac- 
ing between  the  elements,  up  to  a maximum  of 
about  one  half  wavelength.  If  the  half  wave 
elements  are  fed  with  equal  lengths  of  trans- 
mission line,  poled  correctly,  a gain  of  about 
3.3  db  is  produced.  Such  an  antenna  is  shown 
in  figure  13.  By  means  of  a phase  reversing 
switch,  the  two  elements  may  be  operated  out 
of  phase,  producing  a cloverleaf  pattern  with 
slightly  less  maximum  gain. 

A three  element  "ptecut”  array  for  40  meter 
operation  is  shown  in  figure  14.  It  is  fed  di- 
rectly with  300  ohm  "ribbon  line,”  and  may 
be  matched  to  a 52  ohm  coaxial  output  trans- 
mitter by  means  of  a Baiun,  such  as  the  Barker 
& Williamson  3975.  The  antenna  has  a gain  of 
about  3.2  db,  and  a beam  width  at  half-power 
points  of  40  degrees. 


23-6  Broadside  Arrays 

Colinear  elements  may  be  stacked  above  or 
below  another  string  of  colinear  elements  to 
produce  what  is  commonly  called  a broadside 
array.  Such  an  array,  when  horizontal  elements 
are  used,  possesses  vertical  directivity  in 
proportion  to  the  number  of  broadsided  (ver- 
tically stacked)  sections  which  have  been 
used.  Since  broadside  arrays  do  have  good  ver- 
tical directivity  their  use  is  recommended  on 
the  14-Mc.  band  and  on  those  higher  in  fre- 
quency. One  of  the  most  popular  of  simple 
broadside  arrays  is  the  "Lazy  H”  array  of  fig- 
ure 15.  Horizontal  colinear  elements  stacked 
two  above  two  make  up  this  antenna  system 
which  is  highly  recommended  for  work  on  fre- 
quencies above  perhaps  14-Mc.  when  moderate 
gain  without  too  much  directivity  is  desired. 
It  has  high  radiation  resistance  and  a gain  of 
approximately  5.5  db.  The  high  radiation  re- 
sistance results  in  low  voltages  and  a broad 
resonance  curve,  which  permits  use  of  inex- 
pensive insulators  and  enables  the  array  to  be 
used  over  a fairly  wide  range  in  frequency. 
For  dime'nsions,  see  the  stacked  dipole  design 
table. 

Stacked  Vertical  stacking  may  be  applied 

Dipoles  to  strings  of  colinear  elements 

longer  than  two  half  waves.  In 
such  arrays,  the  end  quarter  wave  of  each 
string  of  radiators  usually  is  bent  in  to  meet 


Lt  L1 


© 


Figure  15 

THE  "LAZY  H"  ANTENNA  SYSTEM 

Stacking  the  co/ineor  pairs  gives  both  hori- 
zontal and  vertical  directivity.  As  shown,  the 
array  will  give  about  5.5  db  gain.  Note’thot 
the  array  may  be  fed  either  at  the  center  of 
the  phasing  section  or  at  the  bottom;  if  fed 
at  the  bottom  the  phasing  section  must  be 
twisted  through  180°. 


a similar  bent  quarter  wave  from  the  opposite 
end  radiator.  This  provides  better  balance  and 
better  coupling  between  the  upper  and  lower 
elements  when  the  array  is  current-fed.  Arrays 
of  this  type  ate  shown  in  figure  16,  and  are 
commonly  known  as  curtain  arrays. 

Correct  length  for  the  elements  and  stubs 
can  be  determined  for  any  stacked  dipole  array 
from  the  Stacked-Dipole  Design  Table. 

In  the  sketches  of  figure  16  the  arrowheads 
represent  the  direction  of  current  flow  at  any 
given  instant.  The  dots  on  the  radiators  tepre- 


HANDBOOK 


Broadside  Arrays  469 


1-3  L2  L3 


Figure  16 

THE  STERBA  CURTAIN  ARRAY 

Approximate  directive  gains  a/ong  with  alter- 
native feed  methods  are  shown. 


sent  points  of  maximum  current.  All  arrows 
should  point  in  the  same  direction  in  each  por- 
tion of  the  radiating  sections  of  an  antenna  in 
order  to  provide  a field  in  phase  for  broadside 
radiation.  This  condition  is  satisfied  for  the 
arrays  illustrated  in  figure  16.  Figures  16A 
and  16C  show  simple  methods  of  feeding 
a short  Sterba  curtain,  while  an  alternative 
method  of  feed  is  shown  in  the  higher  gain  an- 
tenna of  figure  16B. 

In  the  case  of  each  of  the  arrays  of  figure 
16,  and  also  the  "Lazy  H”  of  figure  15,  the 
array  may  be  made  unidirectional  and  the  gain 
increased  by  3 db  if  an  exactly  similar  array 
is  constructed  and  placed  approximately  54 
wave  behind  the  driven  array.  A screen  or  mesh 
of  wires  slightly  greater  in  area  than  the  an- 
tenna array  may  be  used  instead  of  an  addi- 
tional array  as  a reflector  to  obtain  a unidirec- 
tional system.  The  spacing  between  the  re- 
flecting wires  may  vary  from  0.05  to  0.1  wave- 
length with  the  spacing  between  the  reflecting 
wires  the  smallest  directly  behind  the  driven 
elements.  The  wires  in  the  untuned  reflecting 
system  should  be  parallel  to  the  radiating  ele- 
ments of  the  array,  and  the  spacing  of  the  com- 
plete reflector  system  should  be  approximately 
0.2  to  0.25  wavelength  behind  the  driven  ele- 
ments. 

On  frequencies  below  perhaps  100  Me.  it 
normally  will  be  impracticable  to  use  a wire- 
screen  reflector  behind  an  antenna  array  such 


as  a Sterba  curtain  or  a "Lazy  H.”  Parasitic 
elements  may  be  used  as  reflectors  or  direc- 
tors, but  parasitic  elements  have  the  disadvan- 
tage that  their  operation  is  selective  with  re- 
spect to  relatively  small  changes  in  frequency. 
Nevertheless,  parasitic  reflectors  for  such  ar- 
rays are  quite  widely  used. 

The  X-Array  In  section  23-5  it  was  shown 
how  two  dipoles  may  be  arranged 
in  phase  to  provide  a power  gain  of  (some)  3 
db.  If  two  such  pairs  of  dipoles  are  stacked 


LAZY-H  AND  STERBA 
(STACKED  DIPOLE)  DESIGN  TABLE 

FREQUENCY 

IN  MC. 

Li 

Ln 

U 

7.0 

68'2" 

70' 

35' 

7,3 

65'10" 

67'6" 

33'9" 

14.0 

34'1" 

35' 

17'6" 

14.2 

33'8" 

34'7" 

17'3" 

14.4 

33'4" 

34'2" 

17' 

21.0 

22'9" 

23'3" 

11 '8" 

21.5 

22'3" 

22'9" 

n'5" 

27.3 

17'7" 

17'10" 

8*11" 

28.0 

17' 

Ml" 

8'9" 

29.0 

16'6" 

17' 

8'6" 

50.0 

9'7" 

9'10" 

4'!!" 

52.0 

9'3" 

9'5" 

4'8" 

54.0 

8'10" 

9'1" 

4'6" 

144.0 

39.8" 

40.5" 

20.3" 

146.0 

39" 

40" 

20" 

148.0 

38.4" 

39.5" 

19.8" 

47  0 High  Frequency  Directive  Antennas 


the  radio 


. L L • 


DIMENSIONS 


lOM.  I5M.  20M. 


L 

I6'3*| 

22' 

32' 10- 

S 

20' 

30' 

40' 

P 

I4'2' 

ei'S' 

2»'4- 

D 

S'C*' 

5'3- 

7'0* 

GAIN  APPROX,  a DB 


-73  A TRANSMISSION  LINE 


Figure  17 

THE  X-ARRAY  FOR  28  MC,  21  MC, 
OR  14  MC. 

The  entire  array  (with  the  exception  of  the 
75-o/im  feed  line)  is  constructed  of  300-ohm 
ribbon  line.  Be  sure  phasing  lines  (P)  are 
poled  correctly^  as  shown. 


in  a vertical  plane  and  properly  phased,  a 
simplified  form  of  in-phase  curtain  is  formed, 
providing  an  overall  gain  of  about  6 db.  Such 
an  array  is  shown  in  figure  17.  In  this  X-array, 
the  four  dipoles  are  all  in  phase,  and  are  fed 
by  four  sections  of  300-ohm  line,  each  one- 
half  wavelength  long,  the  free  ends  of  all  four 
lines  being  connected  in  parallel.  The  feed 
impedance  at  the  junction  of  these  four  lines 
is  about  75  ohms,  and  a length  of  75-ohm 
Twin-Lead  may  be  used  for  the  feedline  to 
the  array. 

An  array  of  this  type  is  quite  small  for  the 
28-Mc.  band,  and  is  not  out  of  the  question 
for  the  21-Mc.  band.  For  best  results,  the  bot- 
tom section  of  the  array  should  be  one-half 
wavelength  above  ground. 

The  Double-Bruce  The  Bruce  Beam  consists 
Array  of  a long  wire  folded  so 

that  vertical  elements 
carry  in-phase  currents  while  the  horizontal 
elements  carry  out  of  phase  currents.  Radia- 
tion from  the  horizontal  sections  is  low  since 
only  a small  current  flows  in  this  part  of  the 
wire,  and  it  is  largely  phased-out.  Since  the 
height  of  the  Bruce  Beam  is  only  one-quarter 
wavelength,  the  gain  per  linear  foot  of  array 
is  quite  iow.  Two  Bruce  Beams  may  be  com- 
bined as  shown  in  figure  18  to  produce  the 
Double  Bruce  array.  A four  section  Double 
Bruce  will  give  a vertically  polarized  emis- 
sion, with  a power  gain  of  5 db  over  a simple 


Figure  18 

THE  DOUBLE-BRUCE  ARRAY  FOR  10, 
IS,  AND  20  METERS 

If  a 600-ohm  feed  line  Is  used,  the  20-meter 
array  will  also  perform  on  10  meters  as  a 
Sterba  curtain,  with  an  approximate  gain  of 
9 db. 


dipole,  and  is  a very  simple  beam  to  construct. 
This  antenna,  like  other  so-called  "broadside” 
arrays,  radiates  maximum  power  at  right  angles 
to  the  plane  of  the  array. 

The  feed  impedance  of  the  Double  Bruce  is 
about  750  ohms.  The  array  may  be  fed  with  a 
one-quarter  wave  stub  made  of  300-ohm  ribbon 
line  and  a feedline  made  of  150-ohm  ribbon 
line.  Alternatively,  the  array  may  be  fed  di- 
rectly with  a wide-spaced  600-ohm  transmis- 
sion line  (figure  18).  The  feedline  should 
be  brought  away  from  the  Double  Bruce  for  a 
short  distance  before  it  drops  downward,  to 
prevent  interaction  between  the  feedline  and 
the  lower  part  of  the  center  phasing  section  of 
the  array.  For  best  results,  the  bottom  sec- 
tions of  the  array  should  be  one-half  wave- 
length above  ground. 

Arrays  such  as  the  X-array  and  the  Double 
Bruce  are  essentially  high  impedance  devices, 
and  exhibit  relatively  broad-band  characteris- 
tics. They  are  less  critical  of  adjustment  than 
a parasitic  array,  and  they  work  well  over  a 
wide  frequency  range  such  as  is  encountered 
on  the  28-29-7  Me.  band. 

The  *'Bi-Square”  Illustrated  in  figure  19  is  a 
Broodside  Array  simple  method  of  feeding  a 
small  broadside  array  first 
described  by  W6BCX  several  years  ago  as  a 
practical  method  of  suspending  an  effective 
array  from  a single  pole.  As  two  arrays  of  this 
type  can  be  supported  at  right  angles  from  a 
single  pole  without  interaction,  it  offers  a 
solution  to  the  problem  of  suspending  two  ar- 
rays in  a restricted  space  with  a minimum  of 
erection  work.  The  free  space  directivity  gain 
is  slightly  less  than  that  of  a Lazy  H,  but  is 


HANDBOOK 


Broadside  Arrays  471 


Figure  19 

THE  "BUSQUARE''  BROADSIDE  ARRAY 

Th/s  b/<//recfionof  array  is  ralafd  to  the 
"Lazy  H,"  and  in  spite  of  the  oblique  e/e« 
menfs,  is  horizontally  polarized.  It  has  slight- 
ly less  gain  and  directivity  than  the  Lazy 
H,  the  free  space  directivity  gain  being  ap- 
proximately 4 db.  Its  chief  advantage  is  the 
fact  that  only  a single  pole  is  required  for 
support,  and  two  such  arrays  may  be  sup- 
ported from  a single  pole  without  interaction 
if  the  planes  of  the  elemertts  are  at  right 
angles.  A 600-ohm  line  may  be  suhstituteJ 
for  the  Twin-Lead,  and  either  operated  as  a 
resonant  line,  or  made  non-resonant  by  the 
incorporation  of  a matching  stub. 


still  worthwhile,  being  approximately  4 db  over 
a half-wave  horizontal  dipole  at  the  same  aver- 
age elevation. 

When  two  Bi-Square  arrays  are  suspended 
at  right  angles  to  each  other  (for  general  cov- 
erage) from  a single  pole,  the  Q sections 
should  be  well  separated  or  else  symmetrically 
arranged  in  the  form  of  a square  (the  diagonal 
conductors  forming  one  Q section)  in  order  to 
minimize  coupling  between  them.  The  same 
applies  to  the  line  if  open  construction  is  used 
instead  of  Twin-Lead,  but  if  Twin-Lead  is 
used  the  coupling  can  be  made  negligible  sim- 
ply by  separating  the  two  Twin-Lead  lines  by 
at  least  two  inches  and  twisting  one  Twin- 
Lead  so  as  to  effect  a transposition  every  foot 
or  so. 


NOTE  : S/0£  LENGTH-  ft' 3"  FOR  Z1  MC. 

17'  7‘  FOR  14  MC. 

ELEMENT  SPACING  98"  FOR  EACH  BAND. 
STUB  LENGTH  APPROX.  IS"  FOR  2 1 MC. 

ZO"  FOR  14  MC. 


Figure  20 

THE  CUBICAL-QUAD  ANTENNA  FOR 
THE  10-METER  BAND 


When  tuned  feeders  are  employed,  the  Bi- 
Square  array  can  be  used  on  half  frequency  as 
an  end-fire  vertically  polarized  array,  giving 
a slight  practical  dx  signal  gain  over  a verti- 
cal half-wave  dipole  at  the  same  height. 

A second  Bi-Square  serving  as  a reflector 
may  be  placed  0.15  wavelength  behind  this  an- 
tenna to  provide  an  overall  gain  of  8.5  db.  The 
reflector  may  be  tuned  by  means  of  a quarter- 
wave  stub  which  has  a moveable  shorting  bar 
at  the  bottom  end.  The  stub  is  used  as  a sub- 
stitute for  the  Q-section,  since  the  reflector 
employs  no  feed  line. 

The  "Cubical-  A smaller  version  of  the  Bi- 
Quad"  Antenna  Square  antenna  is  the  Cubi- 
cal-Quad antenna.  Two  half- 
waves of  wire  are  folded  into  a square  that  is 
one-quarter  wavelength  on  a side,  as  shown  in 
figure  20.  The  array  radiates  a horizontally 
polarized  signal.  A reflector  placed  0.15  wave- 
length behind  the  antenna  provides  an  overall 
gain  of  some  6 db.  A shorted  stub  with  a paral- 
leled tuning  capacitor  is  used  to  resonate  the 
reflector. 

The  Cubical-Quad  is  fed  with  a 150-ohm 
line,  and  should  employ  some  sort  of  antenna 
tuner  at  the  transmitter  eijd  of  the  line  if  a pi- 
network  type  transmitter  is  used.  There  is  a 
small  standing  wave  on  the  line,  and  an  open 


472  High  Frequency  Directive  Antennas 


THE  RADIO 


Figure  21 

THE  "SIX-SHOOTER"  BROADSIDE  ARRAY 


wire  line  should  be  employed  if  the  antenna  is 
used  with  a high  power  transmitter. 

To  tune  the  reflector,  the  hack  of  the  an- 
tenna is  aimed  at  a nearby  field-strength  meter 
and  the  reflector  stub  capacitor  is  adjusted 
for  minimum  received  signal  at  the  operating 
frequency. 

This  antenna  provides  high  gain  for  its  small 
size,  and  is  recommended  for  28-Mc.  work. 
The  elements  may  be  made  of  number  14  en- 
amel wire,  and  the  array  may  be  built  on  a light 
bamboo  or  wood  framework. 

The  "Six-Shooter”  As  a good  compromise  be- 
Broodside  Array  tween  gain,  directivity, 
compactness,  mechanical 
simplicity,  ease  of  adjustment,  and  band  width 
the  array  of  figure  21  is  recommended  for  the 
10  to  30  Me.  range  when  the  additional  array 
width  and  greater  directivity  are  not  obtain- 
able. The  free  space  directivity  gain  is  ap- 
proximately 7.5  db  over  one  element,  and  the 
practical  dx  signal  gain  over  one  element  at 
the  same  average  elevation  is  of  about  the 
same  magnitude  when  the  array  is  sufficiently 
elevated.  To  show  up  to  best  advantage  the 
atray  should  be  elevated  sufficiently  to  put 
the  lower  elements  well  in  the  clear,  and  pre- 
ferably at  least  0.5  wavelength  above  ground. 

The  "Bobtail”  Another  application  of  ver- 

Bidirectionol  deal  orientation  of  the  ra- 

Broodside  Curtain  diating  elements  of  an  ar- 
ray in  order  to  obtain  low- 
angle  radiation  at  the  lower  end  of  the  h-f 
range  with  low  pole  heights  is  illustrated  in 
figure  22.  When  precut  to  the  specified  dimen- 
sions this  single  pattern  array  will  perform 
well  over  the  7-Mc.  amateur  band  or  the  4-Mc. 
amateur  phone  band.  For  the  4-Mc.  band  the 
requited  two  poles  need  be  only  70  feet  high, 
and  the  array  will  provide  a practical  signal 


Figure  22 

"BOBTAIL"  BIDIRECTIONAL  BROAD- 
SIDE CURTAIN  FOR  THE  7-MC.  OR  THE 
4.0.MC.  AMATEUR  BANDS 


This  simple  vertically  polarized  array  pro- 
vides low  angle  radiation  and  response  with 
comparatively  low  pole  heights,  and  is  very 
effective  for  dx  work  on  the  7-Mc.  band  or 
the  4.0-Mc.  phone  band.  Because  of  the 
phase  relationships,  radiation  from  the  hori- 
zontal portion  of  the  antenna  is  effectively 
suppressed.  Very  little  current  flows  in  the 
ground  lead  to  the  coupling  tank;  so  an  elab- 
orate ground  system  is  not  required,  and  the 
length  of  the  ground  lead  Is  not  critical  so 
long  as  If  uses  heavy  wire  and  Is  reason- 
ably short. 


gain  averaging  from  7 to  10  db  over  a horizon- 
tal half-wave  dipole  utilizing  the  same  pole 
height  when  the  path  length  exceeds  2500 
miles. 

The  horizontal  directivity  is  only  moderate, 
the  beam  width  at  the  half  power  points  being 
slightly  greater  than  that  obtained  from  three 
cophased  vertical  radiators  fed  with  equal  cur- 
rents. This  is  explained  by  the  fact  that  the 
current  in  each  of  the  two  outer  radiators  of 
this  array  carries  only  about  half  as  much  cur- 
rent as  the  center,  driven  element.  While  this 
"binomial"  current  distribution  suppresses 
the  end-fire  lobe  that  occurs  when  an  odd  num- 
ber of  parallel  radiators  with  half-wave  spac- 
ing are  fed  equal  currents,  the  array  still  ex- 
hibits some  high-angle  radiation  and  response 
off  the  ends  as  a result  of  imperfect  cancella- 
tion in  the  flat  top  portion.  This  is  not  suffi- 
cient to  affect  the  power  gain  appreciably,  but 
does  degrade  the  discrimination  somewhat. 

A moderate  amount  of  sag  can  be  tolerated 
at  the  center  of  the  flat  top,  where  it  connects 
to  the  driven  vertical  element.  The  poles  and 
antenna  tank  should  be  so  located  with  respect 
to  each  other  that  the  driven  vertical  element 
drops  approximately  straight  down  from  the 
flat  top. 


HANDBOOK 


Endfire  Arrays  473 


Normally  the  antenna  tank  will  be  located 
in  the  same  room  as  the  transmitter,  to  facil- 
itate adjustment  when  changing  frequency.  In 
this  case  it  is  recommended  that  the  link  cou- 
pled tank  be  located  across  the  room  from  the 
transmitter  if  much  power  is  used,  in  order  to 
minimize  r-f  feedback  difficulties  which  might 
occur  as  a result  of  the  asymmetrical  high  im- 
pedance feed.  If  tuning  of  the  antenna  tank 
from  the  transmitter  position  is  desired,  flex- 
ible shafting  can  be  run  from  the  antenna  tank 
condenser  to  a control  knob  at  the  transmitter. 

The  lower  end  of  the  driven  element  is  quite 
"hot”  if  much  power  is  used,  and  the  lead-in 
insulator  should  be  chosen  with  this  in  mind. 
The  ground  connection  need  not  have  very  low 
resistance,  as  the  current  flowing  in  the 
ground  connection  is  comparatively  small.  A 
stake  or  pipe  driven  a few  feet  in  the  ground 
will  suffice.  However,  the  ground  lead  should 
be  of  heavy  wire  and  preferably  the  length 
should  not  exceed  about  10  feet  at  7 Me.  or 
about  20  feet  at  4 Me.  in  order  to  minimize 
reactive  effects  due  to  its  inductance.  If  it  is 
impossible  to  obtain  this  short  a ground  lead, 
a piece  of  screen  or  metal  sheet  about  four 
feet  square  may  be  placed  parallel  to  the  earth 
in  a convenient  location  and  used  as  an  arti- 
ficial ground.  A fairly  high  C/L  ratio  ordinari- 
ly will  be  required  in  the  antenna  tank  in  order 
to  obtain  adequate  coupling  and  loading. 


23-7  End-Fire  Directivity 

By  spacing  two  half-wave  dipoles,  or  colin- 
ear  arrays,  at  a distance  of  from  0.1  to  0.25 
wavelength  and  driving  the  two  180°  out  of 
phase,  directivity  is  obtained  through  the  two 
wires  at  right  angles  to  them.  Hence,  this  type 
of  bidirectional  array  is  called  end  fire.  A bet- 
ter idea  of  end-fire  directivity  can  be  obtained 
by  referring  to  figure  10. 

Remember  that  end- fire  refers  to  the  radia- 
tion with  respect  to  the  two  wires  in  the  array 
rather  than  with  respect  to  the  array  as  a 
whole. 

The  vertical  directivity  of  an  end-fire  bi- 
directional array  which  is  oriented  horizontal- 
ly can  be  increased  by  placing  a similar  end- 
fire  array  a half  wave  below  it,  and  excited  in 
the  same  phase.  Such  an  array  is  a combina- 
tion broadside  and  end- fire  affair. 

Kraus  Flat-Top  A very  effective  bidirectional 
Beom  end-fire  array  is  the  Kraus  or 

8JK  Flat-Top  Beam.  Essen- 
tially, this  antenna  consists  of  two  close- 
spaced dipoles  or  colinear  arrays.  Because  of 
the  close  spacing,  it  is  possible  to  obtain  the 


proper  phase  relationships  in  multi-section 
flat  tops  by  crossing  the  wires  at  the  voltage 
loops,  rather  than  by  resorting  to  phasing 
stubs.  This  greatly  simplifies  the  array.  (See 
figure  23.)  Any  number  of  sections  may  be 
used,  though  the  one-  and  two-section  arrange- 
ments are  the  most  popular.  Little  extra  gain 
is  obtained  by  using  more  than  four  sections, 
and  trouble  from  phase  shift  may  appear. 

A center-fed  single-section  flat-top  beam 
cut  according  to  the  table,  can  be  used  quite 
successfully  on  its  second  harmonic,  the  pat- 
tern being  similar  except  that  it  is  a little 
sharper.  The  single-section  array  can  also  be 
used  on  its  fourth  harmonic  with  some  success, 
though  there  then  will  be  four  cloverleaf  lobes, 
much  the  same  as  with  a full-wave  antenna. 

If  a flat-top  beam  is  to  be  used  on  more  than 
one  band,  tuned  feeders  are  necessary. 

The  radiation  resistance  of  a flat-top  beam 
is  rather  low,  especially  when  only  one  sec- 
tion is  used.  This  means  that  the  voltage  will 
be  high  at  the  voltage  loops.  For  this  reason, 
especially  good  insulators  should  be  used  for 
best  results  in  wet  weather. 

The  exact  lengths  for  the  radiating  elements 
ate  not  especially  critical,  because  slight  de- 
viations from  the  correct  lengths  can  be  com- 
pensated in  the  stub  or  tuned  feeders.  Proper 
stub  adjustment  is  coveted  in  Chapter  Twenty- 
five.  Suitable  radiator  lengths  and  approximate 
stub  dimensions  are  given  in  the  accompany- 
ing design  table. 

Figure  23  shows  top  views  of  eight  types 
of  flat-top  beam  antennas.  The  dimensions  for 
using  these  antennas  on  different  bands  ate 
given  in  the  design  table.  The  7-  and  28-Mc. 
bands  are  divided  into  two  parts,  but  the  di- 
mensions for  either  the  low-  or  high-frequency 
ends  of  these  bands  will  be  satisfactory  for 
use  over  the  entire  band. 

In  any  case,  the  antennas  are  tuned  to  the 
frequency  used,  by  adjusting  the  shorting  wire 
on  the  stub,  or  tuning  the  feeders,  if  no  stub 
is  used.  The  data  in  the  table  may  be  extended 
to  other  bands  or  frequencies  by  applying  the 
proper  factor.  Thus,  for  50  to  52  Me.  operation, 
the  values  for  28  to  29  Me.  are  divided  by  1.8. 

All  of  the  antennas  have  a bidirectional  hori- 
zontal pattern  on  their  fundamental  frequency. 
The  maximum  signal  is  broadside  to  the  flat 
top.  The  single- section  type  has  this  pattern 
on  both  its  fundamental  frequency  and  second 
harmonic.  The  other  types  have  four  main  lobes 
of  radiation  on  the  second  and  higher  harmon- 
ics. The  nominal  gains  of  the  different  types 
over  a half-wave  comparison  antenna  are  as 
follows;  single-section,  4 db;  two-section,  6 
db;  four-section,  8 db. 

The  maximum  spacings  given  make  the 
beams  less  critical  in  their  adjustments.  Up 


474  High  Frequency  Directive  Antennas 


THE  RADIO 


CENTER  FED 


TO  CEKT£R 
or  rLAT  TOP 


MATCHING  STUB 


[- — l3 — 


FIGURE  23 


L4  Hvr  La- 


FLAT.TOP  BEAM  (8JK  ARRAY)  DESIGN  DATA. 


FREQUENCY 

Spac- 

int 

s 

Li 

Li 

U 

1a 

M 

D 

A (Vi) 
approx. 

A(>A) 

approx. 

A (V4) 
approx. 

X 

approx. 

7.0-7. 2 Me. 

T/F 

11' 4' 

34' 

60' 

52'8' 

44' 

8'io"' 

4' 

2& 

60' 

96' 

4' 

1.2-1. i 

X/8 

17'0' 

33'6' 

59' 

SI'S' 

43'!' 

8'8' 

4' 

26' 

59' 

94' 

4' 

14.0-14.4 

X/8 

8'8' 

17' 

30' 

26'4' 

22' 

4'5' 

2' 

13' 

30' 

48' 

2' 

14.0-14.4 

.15X 

10^5' 

ir 

30' 

25'3' 

20' 

5'4' 

2' 

12' 

29' 

47' 

2' 

14.0-14.4 

.20X 

13'11' 

17' 

30' 

22'10' 

7'2' 

2' 

10' 

27' 

45' 

3' 

14.0-14.4 

X/4 

17'4' 

17' 

30' 

20'8' 

8'10' 

2' 

8' 

25' 

43' 

4' 

28.0-29.0 

.15X 

5^2*' 

8'6' 

15' 

12'7' 

10' 

2'8' 

1'6' 

7' 

15' 

24' 

1' 

28.0-29.0 

X/4 

S'S" 

8'6' 

15' 

10'4' 

4'5' 

I'G" 

5' 

13' 

22' 

2' 

29.0-30.0 

.15\ 

s'e' 

8'3' 

14' 6' 

12'2' 

9'8' 

2'7' 

1'6' 

7' 

15' 

23' 

1' 

29.0-30.0 

x/4 

S'A" 

8'3' 

14'6' 

lO'O' 

4'4*' 

1'6' 

5' 

13' 

21' 

2' 

Dimension  chart  for  flat-top  beam  antennas*  The  meanings  of  the  symbols  are  as  follows: 

Li,  Lt  L»  and  Lk,  the  lengths  of  the  sides  of  the  flat-top  sections  as  shown-  Li  is  lertgth 
of  the  sides  of  single-section  center-fed,  Lt  single-section  end-fed  and  2-section  center-fed,  Lt  4-section 
center-fed  and  end-sections  of  4-section  end-fed,  and  Lt  middle  sections  of  4-section  end-fed, 

S,  the  spacing  between  the  flat-top  wires, 

M,  the  wire  length  from  the  outside  to  the  center  of  each  cross-orer, 

D,  the  spacing  lengthwise  between  sections, 

A (Va),  the  approximate  length  for  a quarter-wave  stub, 

A (V2),  the  approximate  length  for  a half-wave  stub, 

A (Va),  the  approximate  length  for  a three-quarter  wave  stub, 

X,  the  approximate  distance  above  the  shorting  wire  of  the  stub  for  the  connection  of  a 600-ohm 
line.  This  distance,  as  given  in  the  table,  is  approximately  correct  only  for  2-section  flat-tops. 

For  single-section  types  it  will  be  smaller  and  for  3-  and  4-section  types  it  will  be  larger. 

The  lengths  given  for  a half-wave  stub  are  applicable  only  to  single-section  center-fed  flat-tops.  To 
be  certain  of  sufficient  stub  length,  it  is  advisable  to  make  the  stub  a foot  or  so  longer  than  shown  in 
the  table,  especially  with  the  end-fed  types.  The  lengths.  A,  are  measured  from  the  point  where  the 
stub  connects  to  the  flat-top. 

Both  the  center  and  end-fed  types  may  be  used  horizontally.  However,  where  a vertical  antenna  is 
desired,  the  flat-tops  can  be  turned  on  end.  In  this  case,  the  end-fed  types  may  be  more  convenient, 
feeding  from  the  lower  end. 


HANDBOOK 


T riplex  Beam  475 


to  one-quarter  wave  spacing  may  be  used  on 
the  fundamental  for  the  one-section  types  and 
also  the  two-section  center-fed,  but  it  is  not 
desirable  to  use  more  than  0.15  wavelength 
spacing  for  the  other  types. 

Although  the  center-fed  type  of  flat-top  gen- 
erally is  to  be  preferred  because  of  its  sym- 
metry, the  end-fed  type  often  is  convenient  or 
desirable.  For  example,  when  a flat-top  beam 
is  used  vertically,  feeding  from  the  lower  end 
is  in  most  cases  more  convenient. 

If  a multisection  flat-top  array  is  end-fed 
instead  of  center-fed,  and  tuned  feeders  are 
used,  stations  off  the  ends  of  the  array  can  be 
worked  by  tying  the  feeders  together  and  work- 
ing the  whole  affair,  feeders  and  all,  as  a long- 
wire  harmonic  antenna.  A single-pole  double- 
throw switch  can  be  used  for  changing  the 
feeders  and  directivity. 

The  Triplex  The  Triplex  beam  is  a modified 
Beom  version  of  the  W8JK  antenna 

which  uses  folded  dipoles  for 
the  half  wave  elements  of  the  array.  The  use 
of  folded  dipoles  results  in  higher  radiation 
resistance  of  the  array,  and  a high  overall  sys- 
tem performance.  Three  wire  dipoles  are  used 
for  the  elements,  and  300-ohm  Twin-Lead  is 


used  for  the  two  phasing  sections.  A recom- 
mended assembly  for  Triplex  beams  for  28  Me., 
21  Me.,  and  14  Me.  is  shown  in  figure  24.  The 
gain  of  a Triplex  beam  is  about  4.5  db  over  a 
dipole. 


23-8  Combination  End-Fire  and 
Broadside  Arrays 

Any  of  the  end-fire  arrays  previously  de- 
scribed may  be  stacked  one  above  the  other  or 
placed  end  to  end  (side  bv  side)  to  give  great- 
er directivity  gain  while  maintaining  a bidi- 
rectional characteristic.  However,  it  must  be 
kept  in  mind  that  to  realize  a worthwhile  in- 
crease in  directivity  and  gain  while  maintain- 
ing a bidirectional  pattern  the  individual  ar- 
rays must  be  spaced  sufficiently  to  reduce  the 
mutual  impedances  to  a negligible  value. 

When  two  flat  top  beams,  for  instance,  are 
placed  one  above  the  other  or  end  to  end,  a 
center  spacing  on  the  order  of  one  wavelength 
is  required  in  order  to  achieve  a worthwhile 
increase  in  gain,  or  approximately  3 db. 


476  High  Frequency  Directive  Antennas 


Thus  it  is  seen  that,  while  maximum  gain 
occurs  with  two  stacked  dipoles  at  a spacing 
of  about  0.7  wavelength  and  the  space  direc- 
tivity gain  is  approximately  5 db  over  one  ele- 
ment under  these  conditions;  the  case  of  two 
flat  top  or  parasitic  arrays  stacked  one  above 
the  other  is  another  story.  Maximum  gain  will 
occur  at  a greater  spacing,  and  the  gain  over 
one  array  will  not  appreciably  exceed  3 db. 

When  two  broadside  curtains  are  placed  one 
ahead  of  the  other  in  end-fire  relationship,  the 
aggregate  mutual  impedance  between  the  two 
curtains  is  such  that  considerable  spacing  is 
required  in  order  to  realize  a gain  approaching 
3 db  (the  requited  spacing  being  a function  of 
the  size  of  the  curtains).  While  it  is  true  that 
a space  directivity  gain  of  approximately  4 db 
can  be  obtained  by  placing  one,  half-wave  di- 
pole an  eighth  wavelength  ahead  of  another 
and  feeding  them  180  degrees  out  of  phase,  a 
gain  of  less  than  1 db  is  obtained  when  the 
same  procedure  is  applied  to  two  large  broad- 
side curtains.  To  obtain  a gain  of  approximate- 
ly 3 db  and  retain  a bidirectional  pattern,  a 
spacing  of  many  wavelengths  is  required  be- 
tween two  large  curtains  placed  one  ahead  of 
the  other. 

A different  situation  exists,  however,  when 
one  driven  curtain  is  placed  ahead  of  an  iden- 
tical one  and  the  two  are  phased  so  as  to  give 
a unidirectional  pattern.  When  a unidirectional 
pattern  is  obtained,  the  gain  over  one  curtain 
will  be  approximately  3 db  regardless  of  the 
spacing.  For  instance,  two  large  curtains 
placed  one  a quarter  wavelength  ahead  of  the 
other  may  have  a space  directivity  gain  of  only 
0.5  db  over  one  curtain  when  the  two  ate  driv- 
en 180  degrees  out  of  phase  to  give  a bidirec- 
tional pattern  (the  type  of  pattern  obtained 
with  a single  curtain).  However,  if  they  are 
driven  in  phase  quadrature  (and  with  equal  cur- 
rents) the  gain  is  approximately  3 db. 

The  directivity  gain  of  a composite  array 
also  can  be  explained  upon  the  basis  of  the 
directivity  patterns  of  the  component  arrays 
alone,  but  it  entails  a rather  complicated  pic- 
ture. It  is  sufficient  for  the  purpose  of  this 
discussion  to  generalize  and  simplify  by  say- 
ing that  the  greater  the  directivity  of  an  end- 
fire  array,  the  farther  an  identical  array  must 
be  spaced  from  it  in  broadside  relationship  to 
obtain  optimum  performance;  and  the  greater 
the  directivity  of  a broadside  array,  the  farther 


an  identical  array  must  be  spaced  from  it  in 
end-fire  relationship  to  obtain  optimum  per- 
formance and  retain  the  bidirectional  charac- 
teristic. 

It  is  important  to  note  that  while  a bidirec- 
tional end-fire  pattern  is  obtained  with  two 
driven  dipoles  when  spaced  anything  under  a 
half  wavelength,  and  while  the  proper  phase 
relationship  is  180  degrees  regardless  of  the 
spacing  for  all  spacings  not  exceeding  one 
half  wavelength,  the  situation  is  different  in 
the  case  of  two  curtains  placed  in  end-fire  re- 
lationship to  give  a bidirectional  pattern.  For 
maximum  gain  at  zero  wave  angle,  the  curtains 
should  be  spaced  an  odd  multiple  of  one  half 
wavelength  and  driven  so  as  to  be  180  degrees 
out  of  phase,  or  spaced  an  even  multiple  of 
one  half  wavelength  and  driven  in  the  same 
phase.  The  optimum  spacing  and  phase  rela- 
tionship will  depend  upon  the  directivity  pat- 
tern of  the  individual  curtains  used  alone,  and 
as  previously  noted  the  optimum  spacing  in- 
creases with  the  size  and  directivity  of  the 
component  arrays. 

A concrete  example  of  a combination  broad- 
side and  end-fire  array  is  two  Lazy  H arrays 
spaced  along  the  direction  of  maximum  radia- 
tion by  a distance  of  four  wavelengths  and  fed 
in  phase.  The  space  directivity  gain  of  such 
an  arrangement  is  slightly  less  than  9 db.  How- 
ever, approximately  the  same  gain  can  be  ob- 
tained by  juxtaposing  the  two  arrays  side  by 
side  or  one  over  the  other  in  the  same  plane, 
so  that  the  two  combine  to  produce,  in  effect, 
one  broadside  curtain  of  twice  the  area.  It  is 
obvious  that  in  most  cases  it  will  be  mote  ex- 
pedient to  increase  the  area  of  a broadside 
array  than  to  resort  to  a combination  of  end- 
fire  and  broadside  directivity.  One  exception, 
of  course,  is  where  two  curtains  are  fed  in 
phase  quadratiure  to  obtain  a unidirectional 
pattern  and  space  directivity  gain  of  approxi- 
mately 3 db  with  a spacing  between  curtains 
as  small  as  one  quarter  wavelength.  Another 
exception  is  where  very  low  angle  radiation  is 
desired  and  the  maximum  pole  height  is  strict- 
ly limited.  The  two  aforementioned  Lazy  H 
arrays  when  placed  in  end-fire  relationship 
will  have  a considerably  lower  radiation  angle 
than  when  placed  side  by  side  if  the  array  ele- 
vation is  low,  and  therefore  may  under  some 
conditions  exhibit  appreciably  more  practical 
signal  gain. 


CHAPTER  TWENTY-FOUR 


V-H-F  and  U-H-F  Antennas 


The  very-high-frequency  or  v-h-f  frequency 
range  is  defined  as  that  range  falling  between 
30  and  300  Me.  The  ultra-high-frequency  or 
u-h-f  range  is  defined  as  falling  between  300 
and  3000  Me.  This  chapter  will  be  devoted  to 
the  design  and  construction  of  antenna  sys- 
tems for  operation  on  the  amateur  50-Mc.,  144- 
Mc.,  235-Mc.,  and  420-Mc.  bands.  Although  the 
basic  principles  of  antenna  operation  ate  the 
same  for  all  frequencies,  the  shorter  physical 
length  of  a wave  in  this  frequency  range  and 
the  differing  modes  of  signal  propagation  make 
it  possible  and  expedient  to  use  antenna  sys- 
tems different  in  design  from  those  used  on 
the  range  from  3 to  30  Me. 


24-1  Antenna  Requirements 

Any  type  of  antenna  system  useable  on  the 
lower  frequencies  may  be  used  in  the  v-h-f  and 
u-h-f  bands.  In  fact,  simple  non-directive  half- 
wave or  quarter-wave  vertical  antennas  are 
very  popular  for  general  transmission  and  re- 
ception from  all  directions,  especially  for 
short-range  work.  But  for  serious  v-h-f  or  u-h-f 
work  the  use  of  some  sort  of  directional  an- 
tenna array  is  a necessity.  In  the  first  place, 
when  the  transmitter  power  is  concentrated  in- 
to a narrow  beam  the  apparent  transmitter  pow- 
er at  the  receiving  station  is  increased  many 
times.  A "billboard”  array  or  a Sterba  curtain 
having  a gain  of  I6  db  will  make  a 25-watt 
transmitter  sound  like  a kilowatt  at  the  other 


station.  Even  a much  simpler  and  smaller  three- 
or  four-element  parasitic  array  having  a gain 
of  7 to  10  db  will  produce  a marked  improve- 
ment in  the  received  signal  at  the  other  sta- 
tion. 

However,  as  all  v-h-f  and  u-h-f  workers 
know,  the  most  important  contribution  of  a 
high-gain  antenna  array  is  in  reception.  If  a 
remote  station  cannot  be  heard  it  obviously  is 
impossible  to  make  contact.  The  limiting  fac- 
tor in  v-h-f  and  u-h-f  reception  i s i n almost 
every  case  the  noise  generated  within  the  re- 
ceiver itself.  Atmospheric  noise  is  almost  non- 
existent and  ignition  interference  can  almost 
invariably  be  reduced  to  a satisfactory  level 
through  the  use  of  an  effective  noise  limiter. 
Even  with  a grounded-grid  or  neutralized  triode 
first  stage  in  the  receiver  the  noise  contribu- 
tion of  the  first  tuned  circuit  in  the  receiver 
will  be  relatively  large.  Hence  it  is  desirable 
to  use  an  antenna  system  which  will  deliver 
the  greatest  signal  voltage  to  the  first  tuned 
circuit  for  a given  field  strength  at  the  receiv- 
ing location. 

Since  the  field  intensity  being  produced  at 
the  receiving  location  by  a remote  transmitting 
station  may  be  assumed  to  be  constant,  the  re- 
ceiving antenna  which  intercepts  the  greatest 
amount  of  wave  front,  assuming  that  the  polari- 
zation and  directivity  of  the  receiving  antenna 
is  proper,  will  be  the  antenna  which  gives  the 
best  received  signal-to-noise  ratio.  An  antenna 
which  has  two  square  wavelengths  effective 
area  will  pick  up  twice  as  much  signal  power 
as  one  which  has  one  square  wavelength  area, 
assuming  the  same  general  type  of  antenna  and 


477 


478  V-H-F  and  U-H-F  Antennas 


the  radio 


that  both  are  directed  at  the  station  being  re- 
ceived. Many  instances  have  been  repotted 
where  a frequency  band  sounded  completely 
dead  with  a simple  dipole  receiving  antenna 
but  when  the  receiver  was  switched  to  a three- 
element  or  larger  array  a considerable  amount 
of  activity  from  80  to  160  miles  distant  was 
heard. 


Angle  of  The  useful  portion  of  the  signal 
Radiation  in  the  v-h-f  and  u-h-f  range  for 
short  or  medium  distance  communi* 
cation  is  that  which  is  radiated  at  a very  low 
angle  with  respect  to  the  surface  of  the  earth; 
essentially  it  is  that  signal  which  is  radiated 
parallel  to  the  surface  of  the  earth.  A vertical 
antenna  transmits  a portion  of  its  radiation  at 
a very  low  angle  and  is  effective  for  this  rea- 
son; its  radiation  is  not  necessarily  effective 
simply  because  it  is  vertically  polarized.  A 
simple  horizontal  dipole  radiates  very  little 
low-angle  energy  and  hence  is  not  a satisfac- 
tory v-h-f  or  u-h-f  radiator.  Directive  arrays 
which  concentrate  a major  portion  of  the  radi- 
ated signal  at  a low  radiation  angle  will  prove 
to  be  effective  radiators  whether  their  signal 
is  horizontally  or  vertically  polarized. 

In  all  cases,  the  radiating  system  for  v-h-f 
and  u-h-f  work  should  be  as  high  and  in  the 
clear  as  possible.  Increasing  the  height  of  the 
antenna  system  will  produce  a very  marked 
improvement  in  the  number  and  strength  of  the 
signals  heard,  regardless  of  the  actual  type 
of  antenna  used. 

Transmissian  Transmission  lines  to  v-h-f  and 
Lines  u-h-f  antenna  systems  may  be 

either  of  the  parallel-conductor 
or  coaxial  conductor  type.  Coaxial  line  is  rec- 
ommended for  short  runs  and  closely  spaced 
open-wire  line  for  longer  runs.  Wave  guides 
may  be  used  under  certain  conditions  for  fre- 
quencies greater  than  perhaps  1500  Me.  but 
their  dimensions  become  excessively  great  for 
frequencies  much  below  this  value.  Non-teson- 
ant  transmission  lines  will  be  found  to  be  con- 
siderably more  efficient  on  these  frequencies 
than  those  of  the  resonant  type.  It  is  wise  to 
to  use  the  very  minimum  length  of  transmission 
line  possible  since  transmission  line  losses 
at  frequencies  above  about  100  Me.  mount  very 
rapidly. 

Open  lines  should  preferably  be  spaced 
closer  than  is  common  for  longer  wavelengths, 
as  6 inches  is  an  appreciable  fraction  of  a 
wavelength  at  2 meters.  Radiation  from  the 
line  will  be  greatly  reduced  if  1-inch  or  1/2- 
inch  spacing  is  used,  rather  than  the  mote  com- 
mon 6-inch  spacing. 

Ordinary  TV-type  300-ohm  ribbon  may  be 
used  on  the  2-metet  band  for  feeder  lengths 


of  about  50  feet  or  less.  For  longer  runs,  either 
the  u-h-f  or  v-h-f  TV  open-wire  lines  may  be 
used  with  good  overall  efficiency.  The  v-h-f 
line  is  satisfactory  for  use  on  the  amateur 
420-Mc.  band. 

Antenna  It  is  recommended  that  the  same 

Chongeaver  antenna  be  used  for  transmitting 
and  receiving  in  the  v-h-f  and 
u-h-f  range.  An  ever-present  problem  in  this 
connection,  however,  is  the  antenna  change- 
over relay.  Reflections  at  the  antenna  change- 
over relay  become  of  increasing  importance 
as  the  frequency  of  transmission  is  increased. 
When  coaxial  cable  is  used  as  the  antenna 
transmission  line,  satisfactory  coaxial  anten- 
na changeover  relays  with  low  reflection  can 
be  used.  One  type  manufactured  by  Advance 
Electric  & Relay  Co.,  Los  Angeles  26,  Calif., 
will  give  a satisfactorily  low  value  of  re- 
flection. 

On  the  235-Mc.  and  420-Mc.  amateur  bands, 
the  size  of  the  antenna  array  becomes  quite 
small,  and  it  is  practical  to  mount  two  identi- 
cal antennas  side  by  side.  One  of  these  an- 
tennas is  used  for  the  transmitter,  and  the 
other  antenna  for  the  receiver.  Separate  trans- 
mission lines  are  used,  and  the  antenna  relay 
may  be  eliminated. 

Effect  of  Feed  A vertical  radiator  for 

System  on  Radiation  general  coverage  u-h-f 
Angle  use  should  be  made 

either  ^ or  wavelength 
long.  Longer  vertical  antennas  do  not  have 
dieir  maximum  radiation  at  right  angles  to  the 
line  of  the  radiator  (unless  co-phased),  and, 
therefore,  are  not  practicable  for  use  where 
greatest  possible  radiation  parallel  to  the 
earth  is  desired. 

Unfortunately,  a feed  system  which  is  not 
perfectly  balanced  and  does  some  radiating, 
not  only  robs  the  antenna  itself  of  that  much 
power,  but  distorts  the  radiation  pattern  of  the 
antenna.  As  a result,  the  pattern  of  a vertical 
radiatot  may  be  so  altered  that  the  radiation 
is  bent  upwards  slightly,  and  the  amount  of 
power  leaving  the  antenna  parallel  to  the 
earth  is  greatly  reduced.  A vertical  half-wave 
radiator  fed  at  the  bottom  by  a quarter-wave 
stub  is  a good  example  of  this;  the  slight 
radiation  from  the  matching  section  decreases 
the  power  radiated  parallel  to  the  earth  by 
nearly  10  db. 

The  only  cure  is  a feed  system  which  does 
not  disturb  the  radiation  pattern  of  the  antenna 
itself.  This  means  that  if  a 2-wire  line  is  used, 
the  current  and  voltages  must  be  exactly  the 
same  (though  180°  out  of  phase)  at  any  point 
on  the  feed  line.  It  means  that  if  a concentric 
feed  line  is  used,  there  should  be  no  current 
flowing  on  the  outside  of  the  outer  conductor. 


HANDBOOK 


Antenna  Polarization  479 


Radiator  Cross  There  is  no  point  in  using 
Section  copper  tubing  for  an  antenna 

on  the  medium  frequencies. 
The  reason  is  that  considerable  tubing  would 
be  required,  and  the  cross  section  still  would 
not  be  a sufficiently  large  fraction  of  a wave- 
length to  improve  the  antenna  bandwidth  char- 
acteristics. At  very  high  and  ultra  high  fre- 
quencies, however,  the  radiator  length  is  so 
short  that  the  expense  of  large  diameter  con- 
ductor is  relatively  small,  even  though  copper 
pipe  of  1 inch  cross  section  is  used.  With  such 
conductors,  the  antenna  will  tune  much  more 
broadly,  and  often  a broad  resonance  charac- 
teristic is  desirable.  This  is  particularly  true 
when  an  antenna  or  array  is  to  be  used  over 
an  entire  amateur  band. 

It  should  be  kept  in  mind  that  with  such 
large  cross  section  radiators,  the  resonant 
length  of  the  radiator  will  he  somewhat  shorter, 
being  only  slightly  greater  than  0.90  of  a half 
wavelength  for  a dipole  when  heavy  copper 
pipe  is  used  above  100  Me. 

Insulation  The  matter  of  insulation  is  of 
prime  importance  at  very  high  fre- 
quencies. Many  insulators  that  have  very  low 
losses  as  high  as  30  Me.  show  up  rather  poor- 
ly at  frequencies  above  100  Me.  Even  the  low 
loss  ceramics  are  none  too  good  where  the  r-f 
voltage  is  high.  One  of  the  best  and  most  prac- 
tical insulators  for  use  at  this  frequency  is 
polystyrene.  It  has  one  disadvantage,  however, 
in  that  it  is  subject  to  fracture  and  to  deforma- 
tion in  the  presence  of  heat. 

It  is  common  practice  to  design  v-h-f  and 
u-h-f  antenna  systems  so  that  the  various  rad- 
iators are  supported  only  at  points  of  relatively 
low  voltage;  the  best  insulation,  obviously,  is 
air.  The  voltages  on  properly  operated  untuned 
feed  lines  ate  not  high,  and  the  question  of 
insulation  is  not  quite  so  important,  though  in- 
sulation still  should  be  of  good  grade. 

Antenna  Commercial  broadcasting  in  the 

Polarization  U.S.A.  for  both  FM  and  tele- 
vision in  the  v-h-f  range  has 
been  standarized  on  horizontal  polarization. 
One  of  the  main  reasons  for  this  standardiza- 
tion is  the  fact  that  ignition  interference  is 
reduced  through  the  use  of  a horizontally  po- 
larized receiving  antenna.  Amateur  practice, 
however,  is  divided  between  horizontal  and 
vertical  polarization  in  the  v-h-f  and  u-h-f 
range.  Mobile  stations  ate  invariably  vertical- 
cally  polarized  due  to  the  physical  limitations 
imposed  by  the  automobile  antenna  installa- 
tion. Most  of  the  stations  doing  intermittent 
or  occasional  work  on  these  frequencies  use  a 
simple  ground-plane  vertical  antenna  for  both 
transmission  and  reception.  However,  those 


TABLE 

OF  WAVELENGTHS 

Fre- 
quency 
in  Me. 

Va  Wove 
Free 
Space 

^/a  Wove 
An- 
tenna 

Vi  Wove 
Free 

Space 

V2  Wove 
An- 
tenna 

50.0 

59.1 

55.5 

118.1 

111.0 

50.5 

58.5 

55.0 

116.9 

109.9 

51.0 

57.9 

54.4 

115.9 

108.8 

51.5 

57.4 

53.9 

114.7 

107.8 

52.0 

56.8 

53.4 

113.5 

106.7 

52.5 

56.3 

52.8 

112.5 

105.7 

53.0 

55.7 

52.4 

111.5 

104.7 

54.0 

54.7 

51.4 

109.5 

102.8 

144 

20.5 

19.2 

41.0 

38.5 

145 

20.4 

19.1 

40.8 

38.3 

146 

20.2 

18.9 

40.4 

38.0 

147 

20.0 

18.8 

40.0 

37.6 

148 

19.9 

18.6 

39.9 

37.2 

235 

12.6 

11.8 

25.2 

23.6 

236 

12.5 

11.8 

25.1 

23.5 

237 

12.5 

11.7 

25.0 

23.5 

238 

12.4 

11.7 

24.9 

23.4 

239 

12.4 

11.6 

24.8 

23.3 

240 

12.3 

11.6 

24.6 

23.2 

420 

7.05 

6.63 

14.1 

13.25 

425 

6.95 

6.55 

13.9 

13.1 

430 

6.88 

6.48 

13.8 

12.95 

All  dimensions  are  in  inches.  Lengths  have  in 
most  coses  been  rounded  off  to  three  significant 
figures.  "Vi^Wove  Free*Space"  column  shown 
obove  should  be  used  with  Lecher  wires  for  fre- 
quency measurement. 

stations  doing  serious  work  and  striving  for 
maximum-range  contacts  on  the  50-Mc.  and 
144-Mc.  bands  almost  invariably  use  horizon- 
tal polarization. 

Experience  has  shown  that  there  is  a great 
attenuation  in  signal  strength  when  using 
crossed  polarization  (transmitting  antenna 
with  one  polarization  and  receiving  antenna 
with  the  other)  for  all  normal  ground-wave  con- 
tacts on  these  bands.  When  contacts  are  be- 
ing made  through  sporadic-E  reflection,  how- 
ever, the  use  of  crossed  polarization  seems  to 
make  no  discernible  difference  in  signal 
strength.  So  the  operator  of  a station  doing 
v-h-f  work  (particularly  on  the  50-Mc.  band) 
is  faced  with  a problem:  If  contacts  are  to  be 
made  with  all  stations  doing  work  on  the  same 
band,  provision  must  be  made  for  operation  on 
both  horizontal  and  vertical  polarization.  This 
problem  has  been  solved  in  many  cases  through 
the  construction  of  an  antenna  array  that  may 
be  revolved  in  the  plane  of  polarization  in  ad- 
dition to  being  capable  of  .rotation  in  the  azi- 
muth plane. 

An  alternate  solution  to  the  problem  which 
involves  less  mechanical  construction  is  sim- 
ply to  install  a good  ground-plane  vertical  an- 
tenna for  all  vertically-polarized  work,  and 
then  to  use  a multi-element  horizontally-polar- 
ized array  for  dx  work. 

24-2  Simple  Horizontally- 
Polarized  Antennas 

Anterma  systems  which  do  not  concentrate 


480  V-H-F  and  U-H-F  Antennas 


THE  RADIO 


TRANSMISSION  LINE 


(D 


© 


Figure  1 

THREE  NONDIRECTIONAL,  HORIZONTALLY  POLARIZED  ANTENNAS 


radiation  at  the  very  low  elevation  angles  are 
not  recommended  for  v-h-f  and  u-h-f  work.  It  is 
for  this  reason  that  the  horizontal  dipole  and 
horizontally-disposed  colineat  arrays  are  gen- 
erally unsuitable  for  work  on  these  frequen- 
cies. Arrays  using  broadside  or  end-fire  ele- 
ments do  concentrate  radiation  at  low  eleva- 
tion angles  and  are  recommended  for  v-h-f 
work.  Arrays  such  as  the  lazy-H,  Sterba  cur- 
tain, flat-top  beam,  and  arrays  with  parasiti- 
cally  excited  elements  are  recommended  for 
this  work.  Dimensions  for  the  first  three  types 
of  arrays  may  be  determined  from  the  data 
given  in  the  previous  chapter,  and  reference 
maybe  made  to  the  Table  of  Wavelengths  given 
in  this  chapter. 

Arrays  using  vertically-stacked  horizontal 
dipoles,  such  as  are  used  by  commercial  tele- 
vision and  FM  stations,  ate  capable  of  giving 
high  gain  without  a sharp  horizontal  radiation 
pattern.  If  sets  of  crossed  dipoles,  as  shown 
in  figure  lA,  are  fed  90“  out  of  phase  the  re- 
sulting system  is  called  a turnstile  antenna. 
The  90“  phase  difference  between  sets  of  di- 
poles may  be  obtained  by  feeding  one  set  of 
dipoles  with  a feed  line  which  is  one-quarter 
wave  longer  than  the  feed  line  to  the  other 
set  of  dipoles.  The  field  strength  broadside  to 
one  of  the  dipoles  is  equal  to  the  field  from 
that  dipole  alone.  The  field  strength  at  a point 
at  any  other  angle  is  equal  to  the  vector  sum 
of  the  fields  from  the  two  dipoles  at  that  an- 
gle. A nearly  circular  horizontal  pattern  is 
produced  by  this  antenna. 

A second  antenna  producing  a uniform,  hori- 
zontally polarized  pattern  is  shown  in  figure 
IB.  This  antenna  employs  three  dipoles  bent 
to  form  a circle.  All  dipoles  are  excited  in 
phase,  and  are  center  fed.  A bazooka  is  in- 
cluded in  the  system  to  prevent  unbalance  in 
the  coaxial  feed  system. 


A third  nondirectional  antenna  is  shown  in 
figure  1C.  This  simple  antenna  is  made  of  two 
half-wave  elements,  of  which  the  end  quarter- 
wavelength  of  each  is  bent  back  90  degrees. 
The  pattern  from  this  antenna  is  very  much 
like  that  of  the  turnstile  antenna.  The  field 
from  the  two  quarter-wave  sections  that  are 
bent  back  are  additive  because  they  are  180 
degrees  out  of  phase  and  are  a half  wave- 
length apart.  The  advantage  of  this  antenna  is 
the  simplicity  of  its  feed  system  and  con- 
struction. 

24-3  Simple  Vertical-Polarized 
Antennas 

For  general  coverage  with  a single  antenna, 
a single  vertical  radiator  is  commonly  em- 
ployed. A two-wire  open  transmission  line  is 
not  suitable  for  use  with  this  type  antenna, 
and  coaxial  polyethylene  feed  line  such  as 
RG-8/U  is  to  be  recommended.  Three  practical 
methods  of  feeding  the  radiator  with  concen- 
tric line,  with  a minimum  of  current  induced 
in  the  outside  of  the  line,  are  shown  in  figure 
2.  Antenna  (A)  is  known  as  the  sleeve  anten- 
na, the  lower  half  of  the  radiator  being  a large 
piece  of  pipe  up  through  which  the  concentric 
feed  line  is  run.  At  (B)  is  shown  the  ground- 
plane  vertical,  and  at  (C)  a modification  of 
this  latter  antenna. 

The  radiation  resistance  of  the  ground- 
plane  vertical  is  approximately  30  ohms,  which 
is  not  a standard  impedance  for  coaxial  line. 
To  obtain  a good  match,  the  first  quarter  wave- 
length of  feeder  may  be  of  52  ohms  surge  im- 
pedance, and  the  remainder  of  the  line  of  ap- 
proximately 75  ohms  impedance.  Thus,  the 
first  quarter-wave  section  of  line  is  used  as  a 


HANDBOOK 


Vertically  Polarized  Arrays  481 


Figure  2 

THREE  VERTICALLY-POLARIZED 
LOW-ANGLE  RADIATORS 

Shown  at  (A)  is  the  * sleeve'*  or  " hypoder- 
mic"  type  of  radiator.  At  (B)  is  shown  the 
ground-plane  vertical,  and  ( C)  shows  a modi- 
fication of  this  antenna  system  which  in- 
creases the  feed-point  impedance  to  a value 
such  that  the  system  may  be  fed  directly 
from  a coaxial  line  with  no  standing  waves 
on  the  feed  line. 


matching  transformer,  and  a good  match  is 
obtained. 

In  actual  practice  the  antenna  would  con- 
sist of  a quartet-wave  rod,  mounted  by  means 
of  insulators  atop  a pole  or  pipe  mast.  Elab- 
orate insulation  is  not  requited,  as  the  voltage 
at  the  lower  end  of  the  quarter-wave  radiator 
is  very  low.  Self-supporting  rods  from  0.25  to 
0.28  wavelength  would  be  extended  out,  as  in 
the  illustration,  and  connected  together.  As 
the  point  of  connection  is  effectively  at  ground 
potential,  no  insulation  is  required;  the  hori- 
zontal rods  may  be  bolted  directly  to  the  sup- 
porting pole  or  mast,  even  if  of  metal.  The  co- 
axial line  should  be  of  the  low  loss  type  es- 
pecially designed  for  v-h-f  use.  The  outside 
connects  to  the  junction  of  the  radials,  and 
the  inside  to  the  bottom  end  of  the  vertical 
radiator.  An  antenna  of  this  type  is  moderately 
simple  to  construct  and  will  give  a good  ac- 
count of  itself  when  fed  at  the  lower  end  of  the 
radiator  directly  by  the  52-ohm  RG-8/U  co- 
axial cable.  Theoretically  the  standing-wave 
ratio  will  be  approximately  1.5-to-l  but  in 
practice  this  moderate  s-w-r  produces  no 
deleterious  effects,  even  on  coaxial  cable. 

The  modification  shown  in  figure  2C  permits 
matching  to  a standard  50-  or  70-ohm  flexible 
coaxial  cable  without  a linear  transformer.  If 
the  lower  rods  hug  the  lime  and  supporting  mast 


rather  closely,  the  feed-point  impedance  is 
about  70  ohms.  If  they  are  bent  out  to  form  an 
angle  of  about  30°  with  the  support  pipe  the 
impedance  is  about  50  ohms. 

The  number  of  radial  legs  used  in  a ground- 
plane  antenna  of  either  type  has  an  important 
effect  on  the  feed-point  impedance  and  upon 
the  radiation  characteristics  of  the  antenna 
system.  Experiment  has  shown  that  three  radi- 
als is  the  minimum  number  that  should  be 
used,  and  that  increasing  the  number  of  radi- 
als above  six  adds  substantially  nothing  to  the 
effectiveness  of  the  antenna  and  has  no  effect 
on  the  feed-point  impedance.  Experiment  has 
shown,  however,  that  the  radials  should  be 
slightly  longer  than  one-quarter  wave  for  best 
results.  A length  of  0.28  wavelength  has  been 
shown  to  be  the  optimum  value.  This  means 
that  the  radials  for  a 50-Mc.  ground-plane  ver- 
tical antenna  should  be  65  " in  length. 

Double  Skeleton  The  bandwidth  of  the  anten- 
Cone  Antenno  na  of  figure  2C  can  be  in- 
creased considerably  by  sub- 
stituting several  space-tapered  rods  for  the 
single  radiating  element,  so  that  the  "radia- 
tor” and  skirt  are  similar.  If  a sufficient  num- 
ber of  rods  are  used  in  the  skeleton  cones  and 
the  angle  of  revolution  is  optimized  for  the 
particular  type  of  feed  line  used,  this  antenna 
exhibits  a very  low  SWR  over  a 2 to  1 frequen- 
cy range.  Such  an  arrangement  is  illustrated 
schematically  in  figure  3. 

A Nondirectionol  Half-wave  elements  may  be 
Vertical  Array  stacked  in  the  vertical  plane 
to  provide  a non-directional 
pattern  with  good  horizontal  gain.  An  array 
made  up  of  four  half-wave  vertical  elements 
is  shown  in  figure  4A.  This  antenna  provides 
a circular  pattern  with  a gain  of  about  4.5  db 
over  a vertical  dipole.  It  may  be  fed  with 
300-ohm  TV-type  line.  The  feedline  should  be 
conducted  in  such  a way  that  the  vertical  por- 
tion of  the  line  is  at  least  one-half  wavelength 
away  from  the  vertical  antenna  elements.  A 
suitable  mechanical  assembly  is  shown  in  fig- 
ure 4B  for  the  144-Mc.  and  235-Mc.  amateur 
bands. 

24-4  The  Discone  Antenna 


The  Discone  antenna  is  a vertically  polar- 
ized omnidirectional  radiator  which  has  very 
broad  band  characteristics  and  permits  a sim- 
ple, rugged  structure.  This  antenna  presents  a 
substantially  uniform  feed-point  impedance, 
suitable  for  direct  connection  of  a coaxial 
line,  over  a range  of  several  octaves.  Alsq, 
the  vertical  pattern  is  suitable  for  ground-wave 


482  V-H-F  and  U-H-F  Antennas 


THE  RADIO 


Figure  3 

THE  DOUBLE  SKELETON  CONE 
ANTENNA 

A skeleton  cone  has  been  substituted  for  the 
single  element  radiator  of  figure  2C.  This 
greatly  increases  the  bandwidth.  If  at  least 
10  elements  are  used  for  each  skeleton  cone 
and  the  angle  of  revolution  and  element 
length  are  optimized,  a low  SVfR  can  be  ob* 
talned  over  a frequency  range  of  at  least  two 
octaves.  To  obtain  this  order  of  bandwidth, 
the  element  length  L should  be  approximate, 
ly  0.2  wavelength  at  the  lower  frequency  end 
of  the  band,  and  the  angle  of  revolution  opti* 
mixed  for  the  lowest  maximum  VSWR  within 
the  frequency  range  to  be  covered.  A greater 
improvement  in  the  impedance~frequency 
characteristic  can  be  achieved  by  adding 
elements  than  by  increasing  the  diameter  of 
the  elements.  With  only  3 elements  per 
"cone**  and  a much  smaller  angle  of  revo- 
lution  a low  SWR  can  be  obtained  over  a fre- 
quency ronge  of  approximately  1.3  to  1.0 
when  the  element  lengths  are  optimized. 


work  over  several  octaves,  the  gain  varying 
only  slightly  over  a very  wide  frequency  range. 

Conunercial  versions  of  the  Discone  anten- 
na for  various  applications  are  manufactured 
by  the  Federal  Telephone  and  Radio  Corpora- 
tion. A Discone  type  antenna  for  amateur  work 
can  be  fabricated  from  inexpensive  materials 
with  ordinary  hand  tools. 

A Discone  antenna  suitable  for  multi-band 
amateur  work  in  the  v-h/u-h-f  range  is  shown 
schematically  in  figure  5A.  The  distance  D 
should  be  made  approximately  equal  to  a free- 
space  quarter  wavelength  at  the  lowest  oper- 


NONDIRECTIONAL  ARRAYS  FOR  144  MC. 
AND  235  MC. 


On  right  is  shown  two  band  installation.  The 
whole  system  may  easily  be  dissembled  and 
carried  on  a ski-rack  atop  a car  for  portable 
use. 


ating  frequency.  The  antenna  then  will  per- 
form well  over  a frequency  range  of  at  least 
8 to  1.  At  certain  frequencies  within  this 
range  the  vertical  pattern  will  tend  to  "lift” 
slightly,  causing  a slight  reduction  in  gain  at 
zero  angular  elevation,  but  the  reduction  is 
very  slight. 

Below  the  frequency  at  which  the  slant 
height  of  the  conical  skirt  is  equal  to  a ftee- 
space  quarter  wavelength  the  standing-wave 
ratio  starts  to  climb,  and  below  a frequency 
approximately  20  per  cent  lower  than  this  the 
standing-wave  ratio  climbs  very  rapidly.  This 
is  termed  the  cut  ojj  frequency  of  the  antenna. 
By  making  the  slant  height  approximately  equal 
to  a free-space  quarter  wavelength  at  the  low- 
est frequency  employed  (refer  to  chart),  a 


HANDBOOK 


Discone  Antenna  483 


h 0.7  D 


RADIATOR 

This  antenna  system  radiates  a vertically 
polarized  wave  over  a very  wide  frequency 
range.  The  disc”  may  be  made  of  solid 
metal  sheet,  a group  of  radials,  or  wire 
screen;  the  cone”  may  best  be  constructed 
by  forming  a sheet  of  thin  aluminum.  A sin- 
gle antenna  may  be  used  for  operation  on  the 
50,  144,  and  220  Me.  amateur  bands.  The 
dimension  D is  determined  by  the  lowest  fre* 
quency  to  be  employed,  and  is  given  in  the 
chart  of  figure  SB, 


VSWRof  less  than  1.5  will  be  obtained  through- 
out the  operating  range  of  the  antenna. 

The  Discone  antenna  may  be  considered 
as  a cross  between  an  electromagnetic  horn 
and  an  inverted  ground  plane  unipole  antenna. 
It  looks  to  the  feed  line  like  a properly  termi- 
nated high-pass  filter. 

Construction  Details  The  top  disk  and  the 
conical  skirt  may  be 
fabricated  either  from  sheet  metal,  screen  (such 
as  "hardware  cloth”),  or  12  or  mote  "spine” 
radials.  If  screen  is  used  a supporting  frame- 
work of  rod  or  tubing  will  be  necessary  for 
mechanical  strength  except  at  the  higher  fre- 
quencies., If  spines  are  used,  they  should  be 
terminated  on  a stiff  ring  for  mechanical 
strength  except  at  the  higher  frequencies. 

The  top  disk  is  supported  by  means  of  three 
insulating  pillars  fastened  to  the  skirt.  Either 
polystyrene  or  low-loss  ceramic  is  suitable  for 
the  purpose.  The  apex  of  the  conical  skirt  is 
grounded  to  the  supporting  mast  and  to  the 
outer  conductor  of  the  coaxial  line.  The  line 
is  run  down  through  the  supporting  mast.  An 

alternative  arrangement,  one  suitable  for  cer- 
tain mobile  applications,  is  to  fasten  the  base 


Figure  SB 

DESIGN  CHART  FOR  THE  ''DISCONE " 
ANTENNA 


of  the  skirt  directly  to  an  effective  ground 
plane  such  as  the  top  of  an  automobile. 


24-5  Helical  Beam 

Antennas 


Most  v-h-f  and  u-h-f  antennas  ate  either  ver- 
tically polarized  or  horizontally  polarized 
(plane  polarization).  However,  circularly  po- 
larized antennas  have  interesting  characteris- 
tics which  may  be  useful  for  certain  applica- 
tions. The  installation  of  such  an  antenna  can 
effectively  solve  the  problem  of  horizontal  vs. 
vertical  polarization. 

A circularly  polarized  wave  has  its  energy 
divided  equally  between  a vertically  polarized 
component  and  a horizontally  polarized  com- 
ponent, the  two  being  90  degrees  out  of  phase. 
The  circularly  polarized  wave  may  be  either 
"left  handed”  or  "right  handed,”  depending 
upon  whether  the  vertically  polarized  compo- 
nent leads  or  lags  the  horizontal  component. 

A circularly  polarized  antenna  will  respond 
to  any  plane  polarized  wave  whether  horizon- 
tally polarized,  vertically  polarized,  or  diag- 
onally polarized.  Also,  a circular  polarized 
wave  can  be  received  on  a plane  polarized  an- 
tenna, regardless  of  the  polarization  of  the 
latter.  When  using  circularly  polarized  anten- 
nas at  both  ends  of  the  circuit,  however,  both 
must  be  left  handed  or  both  must  be  right 
handed.  This  offers  some  interesting  possi- 
bilities with  regard  to  reduction  of  QRM.  At 


484  V-H-F  and  U-H-F  Antennas 


THE  RADIO 


CONDUCTOR  01  A.  = APPROX-  0.17  A 
A = WAVELENGTH  IN  FREE  SPACE 

Figure  6 

THE  "HELICAL  BEAM"  ANTENNA 

This  type  of  directional  arienna  system 
gives  excellent  performance  over  a frequency 
range  of  1.7  to  7.8  to  1.  Its  dimensions  are 
such  that  it  ordinarily  is  not  practicable, 
however,  for  use  as  a rotatable  orray  on  fre* 
quencies  below  about  100  Me.  The  center 
conductor  of  the  feed  line  should  pass 
through  the  ground  screen  for  connection  to 
the  feed  point.  The  outer  conductor  of  the 
coaxial  line  should  be  grounded  to  the 
ground  screen. 


the  time  of  writing,  there  has  been  no  stand- 
ardization of  the  "twist”  for  general  amateur 
work. 

Perhaps  the  simplest  antenna  configuration 
for  a directional  beam  antenna  having  circular 
polarization  is  the  helical  beam  popularized 
by  Dr.  John  Kraus,  W8JK.  The  antenna  con- 
sists simply  of  a helix  working  against  a 
ground  plane  and  fed  with  coaxial  line.  In  the 
u-h-f  and  the  upper  v-h-f  range  the  physical 
dimensions  ate  sufficiently  small  to  permit 
construction  of  a rotatable  structure  without 
much  difficulty. 

When  the  dimensions  are  optimized,  the 
characteristics  of  the  helical  beam  antenna 
ate  such  as  to  qualify  it  as  a broad  band  an- 
tenna. An  optimized  helical  beam  shows  little 
variation  in  the  pattern  of  the  main  lobe  and 
a fairly  uniform  feed  point  impedance  averag- 
ing approximately  125  ohms  over  a frequency 
range  of  as  much  as  1.7  to  1.  The  direction  of 
"electrical  twist”  (right  or  left  handed)  de- 
pends upon  the  direction  in  which  the  helix  is 
wound. 

A six-turn  helical  beam  is  shown  schemati- 
cally in  figure  6.  The  dimensions  shown  will 
give  good  performance  over  a frequency  range 
of  plus  or  minus  20  per  cent  of  the  design  fre- 
quency. This  means  that  the  dimensions  are 
not  especially  critical  when  the  array  is  to  be 


used  at  a single  frequency  or  over  a narrow 

band  of  frequencies,  such  as  an  amateur  band. 
At  the  design  frequency  the  beam  width  is 
about  50  degrees  and  the  power  gain  about  12 
db,  referred  to  a non-directional  circularly  po- 
larized antenna. 

The  Ground  Screen  For  the  frequency  range 
100  to  500  Me.  a suitable 
ground  screen  can  be  made  from  "chicken 
wire”  poultry  netting  of  1-inch  mesh,  fastened 
to  a round  or  square  frame  of  either  metal  or 
wood.  The  netting  should  be  of  the  type  that 
is  galvanized  after  weaving.  A small,  sheet 
metal  ground  plate  of  diameter  equal  to  ap- 
proximately D/2  should  be  centered  on  the 
screen  and  soldered  to  it.  Tin,  galvanized 
iron,  or  sheet  copper,  is  suitable.  The  outer 
conductor  of  the  RG-63/U  (125  ohm)  coax  is 
connected  to  this  plate,  and  the  inner  conduc- 
tor contacts  the  helix  through  a hole  in  the 
center  of  the  plate.  The  end  of  the  coax  should 
be  taped  with  Scotch  electrical  tape  to  keep 
water  out. 

The  Helix  It  should  be  noted  that  the  beam 
proper  consists  of  six  full  turns. 
The  start  of  the  helix  is  spaced  a distance  of 
S/2  from  the  ground  screen,  and  the  conductor 
goes  directly  from  the  center  of  the  ground 
screen  to  the  start  of  the  helix. 

Aluminum  tubing  in  the  "SO”  (soft)  grade 
is  suitable  for  the  helix.  Alternatively,  lengths 
of  the  relatively  soft  aluminum  electrical  con- 
duit may  be  used.  In  the  v-h-f  range  it  will  be 
necessary  to  support  the  helix  on  either  two 
or  four  wooden  longerons  in  order  to  achieve 
sufficient  strength.  The  longerons  should  be 
of  as  small  cross  section  as  will  provide  suf- 
ficient rigidity,  and  should  be  given  several 
coats  of  varnish.  The  ground  plane  butts 
against  the  longerons  and  the  whole  assembly 
is  supported  from  the  balance  point  if  it  is  to 
be  rotated. 

Aluminum  tubing  in  the  larger  diameters  ordi- 
narily is  not  readily  available  in  lengths  great- 
er than  12  feet.  In  this  case  several  lengths 
can  be  spliced  by  means  of  short  telescoping 
sections  and  sheet  metal  screws. 

The  tubing  is  close  wound  on  a drum  and 
then  spaced  to  give  the  specified  pitch.  Note 
that  the  length  of  one  complete  turn  when 
spaced  is  somewhat  greater  than  the  circumfer- 
ence of  a circle  having  the  diameter  D. 

Broad-Band  A highly  useful  v-h-f  helical 

144  to  225  Me.  beam  which  will  receive  sig- 
Helical  Beam  nals  with  good  gain  over  the 
complete  frequency  range  from 
144  through  225  Me.  may  be  constructed  by 
using  the  following  dimensions  (180  Me.  de- 
sign center): 


HANDBOOK 


Helical  Beam  Antenna  485 


D 22  in. 

S 16M  in. 

G 53  in. 

Tubing  o.d 1 in. 

The  D and  S dimensions  are  to  the  center  of 

the  tubing.  These  dimensions  must  be  held 
rather  closely,  since  the  range  from  144  through 
225  Me.  represents  just  about  the  practical 
limit  of  coverage  of  this  type  of  antenna  sys- 
tem. 

High-Band  Note  that  an  array  constructed 

TV  Coverage  with  the  above  dimensions  will 
give  unusually  good  high-band 
TV  reception  in  addition  to  coveting  the  144- 
Mc.  and  220-Mc.  amateur  bands  and  the  taxi 
and  police  services. 

On  the  144-Mc.  band  the  beam  width  is  ap- 
proximately 60  degrees  to  the  half-power 
points,  while  the  power  gain  is  approximately 
11  db  over  a non-directional  circularly  polar- 
ized antenna.  For  high-band  TV  coverage  the 
gain  will  be  12  to  14  db,  with  a beam  width 
of  about  50  degrees,  and  on  the  220-Mc.  ama- 
teur band  the  beam  width  will  be  about  40  de- 
grees with  a powergain  of  approximately  15  db. 

The  antenna  system  will  receive  vertically 
polarized  or  horizontally  polarized  signals 
with  equal  gain  over  its  entire  frequency  range. 
Conversely,  it  will  transmit  signals  over  the 
same  range,  which  then  can  be  received  with 
equal  strength  on  either  horizontally  polarized 
or  vertically  polarized  receiving  antennas. 
The  standing-wave  ratio  will  be  very  low  over 
the  complete  frequency  range  if  RG-63/U  co- 
axial feed  line  is  used. 


24-6  The  Corner-Reflector 
and  Horn-Type  Antennas 

The  corner-reflector  antenna  is  a good  direc- 
tional radiator  for  the  v-h-f  and  u-h-f  region. 
The  antenna  may  be  used  with  the  radiating 
element  vertical,  in  which  case  the  directivity 
is  in  the  horizontal  or  azimuth  plane,  or  the 
system  may  be  used  with  the  driven  element 


Figure  7 

CONSTRUCTION  OF  THE  "CORNER 
REFLECTOR"  ANTENNA 

Such  on  antenna  is  capable  af  giving  high 
gain  with  a minimum  of  complexity  in  the 
radiating  system.  It  may  be  used  either  with 
horizontal  or  vertical  polarization.  Design 
data  for  the  antenna  is  given  In  the  Corner^ 
Reflector  Design  Table. 


horizontal  in  which  case  the  radiation  is  hori- 
zontally polarized  and  most  of  the  directivity 
is  in  the  vertical  plane.  With  the  antenna  used 
as  a horizontally  polarized  radiating  system 
the  array  is  a very  good  low-angle  beam  array 
although  the  nose  of  the  horizontal  pattern  is 
still  quite  sharp.  When  the  radiator  is  oriented 
vertically  the  corner  reflector  operates  very 
satisfactorily  as  a direction-finding  antenna. 

Design  data  for  the  corner-reflector  antenna 
is  given  in  figure  7 and  in  the  chart  Corner- 
Reflector  Design  Data.  The  planes  which  make 
up  the  reflecting  corner  may  be  made  of  solid 
sheets  of  copper  or  aluminum  for  the  u-h-f 
bands,  although  spaced  wires  with  the  ends 
soldered  together  at  top  and  bottom  may  be 
used  as  the  reflector  on  the  lower  frequencies. 


CORNER-REFLECTOR  DESIGN  DATA 


Corner 

Angle 

Freq. 

Band,  Me. 

R 

S 

H 

A 

L 

G 

Feed 

Imped. 

Approx. 
Gain,  db 

90 

50 

110" 

82" 

140" 

200" 

230" 

18" 

72 

10 

60 

50 

110" 

115" 

140" 

230" 

230" 

18" 

70 

12 

60 

144 

38" 

40" 

48" 

100" 

100" 

5" 

70 

12 

60 

220 

24.5" 

25" 

30" 

72" 

72" 

3" 

70 

12 

60 

420 

13" 

14" 

18" 

36" 

36" 

screen 

70 

12 

NOTE:  Refer  to  figure  ^ for  construction  of  corner-reflector  antenna. 


486  V-H-F  and  U-H-F  Antennas 


THE  RADIO 


(D  VHF  HORIZONTALLY  POLARIZED  HORN 

Figure  8 

TWO  TYPES  OF  HORN  ANTENNAS 
The  ' two  sided  born  of  Figure  8B  may  be 
fed  by  means  of  an  open-wire  transmission 
line. 


Copper  screen  may  also  be  used  for  the  re- 
flecting planes. 

The  values  of  spacing  given  in  the  corner- 
reflector  chart  have  been  chosen  such  that  the 
center  impedance  of  the  driven  element  would 
be  approximately  70  ohms.  This  means  that 
the  element  may  be  fed  directly  with  70-ohm 
coaxial  line,  or  a quarter-wave  matching  trans- 
former such  as  a "Q”  section  may  be  used  to 
provide  an  impedance  match  between  the  cen- 
ter-impedance of  the  element  and  a 460-ohm 
line  constructed  of  no.  12  wire  spaced  2 inches. 

In  many  v-h-f  antenna  systems,  waveguide 
transmission  lines  are  terminated  by  pyramidal 
horn  antennas.  These  horn  antennas  (figure 
8A)  will  transmit  and  receive  either  horizon- 
tally or  vertically  polarized  waves.  The  use  of 
waveguides  at  144  Me.  and  235  Me.,  however, 
is  out  of  the  question  because  of  the  relatively 
large  dimensions  needed  for  a waveguide  oper- 
ating at  these  low  frequencies. 

A modified  type  of  horn  antenna  may  still  be 
used  on  these  frequencies,  since  only  one  par- 
ticular plane  of  polarization  is  of  interest  to 
the  amateur.  In  this  case,  the  horn  antenna 
can  be  simplified  to  two  triangular  sides  of 
the  pyramidal  horn.  When  these  two  sides  are 
insulated  from  each  other,  direct  excitation  at 
the  apex  of  the  horn  by  a two-wire  transmission 
line  is  possible. 

In  a normal  pyramidal  horn,  all  four  triangu- 
lar sides  are  covered  with  conducting  material, 
but  when  horizontal  polarization  alone  is  of 
interest  (as  in  amateur  work)  only  the  vertical 
areas  of  the  horn  need  be  used.  If  vertical  po- 
larization is  required,  only  the  horizontal  areas 


0 

Za-a 

GAIN  (db) 

X 

T 

400 

3 

X 

420 

9 

2X 

390 

15 

Figure  9 

THE  60°  HORN  ANTENNA  FOR  USE  ON 
FREQUENCIES  ABOVE  144  MC. 


of  the  horn  are  employed.  In  either  case,  the 
system  is  unidirectional,  away  from  the  apex 
of  the  horn.  A typical  horn  of  this  type  is  shown 
in  figure  8B.  The  two  metallic  sides  of  the 
horn  are  insulated  from  each  other,  and  the 
sides  of  the  horn  are  made  of  small  mesh 
"chicken  wire”  or  copper  window  screening. 

A pyramidal  horn  is  essentially  a high-pass 
device  whose  low  frequency  cut-off  is  reached 
when  a side  of  the  horn  is  wavelength.  It 
will  work  up  to  infinitely  high  frequencies, 
the  gain  of  the  horn  increasing  by  6 db  every 
time  the  operating  frequency  is  doubled.  The 
power  gain  of  such  a horn  compared  to  a 
wave  dipole  at  frequencies  higher  than  cut- 


Power  gain  (db)  = 

where  A is  the  frontal  area  of  the  mouth  of  the 
horn.  For  the  60  degree  horn  shown  in  figure 
8B  the  formula  simplifies  to: 

Power  gain  (db)  = 8.4  D“,  when  D is  ex- 
pressed in  terms  of  wavelength 

When  D is  equal  to  one  wavelength,  the  pow- 
er gain  of  the  horn  is  approximately  9 db.  The 
gain  and  feed  point  impedance  of  the  60  de- 
gree horn  are  shown  in  figure  9.  A 450  ohm 
open  wire  TV-type  line  may  be  used  to  feed 
the  horn. 


24-7  VHF  Horizontal 

Rhombic  Antenna 

For  v-h-f  transmission  and  reception  in  a 
fixed  direction,  a horizontal  rhombic  permits 


HANDBOOK 


VHF  Rhombic  487 


SIDE  LENGTH,  S 


Figure  10 

V-H-F  RHOMBIC  ANTENNA  DESIGN 
CHART 

The  optimum  tilt  angle  (see  figure  it)  for 
sero-angle  radiation  depends  upon  the 
length  of  the  sides. 


10  to  16  db  gain  with  a simpler  construction 
than  does  a phased  dipole  array,  and  has  the 
further  advantage  of  being  use^l  over  a wide 
frequency  range. 

Except  at  the  upper  end  of  the  v-h-f  range 
a rhombic  array  having  a worthwhile  gain  is 
too  large  to  be  rotated.  However,  in  locations 
75  to  150  miles  from  a large  metropolitan  area 
a rhombic  array  is  ideally  suited  for  working 
into  the  city  on  extended  (horizontally  polar- 
ized) ground-wave  while  at  the  same  time  mak- 
ing an  ideal  antenna  for  TV  reception. 

The  useful  frequency  range  of  a v-h-f  rhom- 
bic array  is  about  2 to  1,  or  about  plus  40%  and 
minus  30%  from  the  design  frequency.  This 
coverage  is  somewhat  less  than  that  of  a high- 
frequency  rhombic  used  for  sky-wave  communi- 
cation. For  ground-wave  transmission  or  recep- 
tion the  only  effective  vertical  angle  is  that 
of  the  horizon,  and  a frequency  range  greater 
than  2 to  1 cannot  be  covered  with  a rhombic 
array  without  an  excessive  change  in  the  ver- 
tical angle  of  maximum  radiation  or  response. 

The  dimensions  of  a v-h-f  rhombic  array  are 
determined  from  the  design  frequency  and  fig- 
ure 10,  which  shows  the  proper  tilt  angle  (see 
figure  11)  for  a given  leg  length.  The  gain  of 
a rhombic  array  increases  with  leg  length. 
There  is  not  much  point  in  constructing  a v-h-f 
rhombic  array  with  legs  shorter  than  about  4 
wavelengths,  and  the  beam  width  begins  to  be- 
come excessively  sharp  for  leg  lengths  greater 
than  about  8 wavelengths.  A leg  length  of  6 
wavelengths  is  a good  compromise  between 
beam  width  and  gain. 

The  tilt  angle  given  in  figure  10  is  based 
upon  a wave  angle  of  zero  degrees.  For  leg 
lengths  of  4 wavelengths  or  longer,  it  will  be 


TOPVIEW 


Figure  1 1 

V-H-F  RHOMBIC  ANTENNA 
CONSTRUCTION 


necessary  to  elongate  the  array  a few  per  cent 
(pulling  in  the  sides  slightly)  if  the  horizon 
elevation  exceeds  about  3 degrees. 

Table  I gives  dimensions  for  two  dual  pur- 
pose rhombic  arrays.  One  covers  the  6-meter 
amateur  band  and  the  "low”  television  band. 
The  other  covers  the  2-metet  amateur  band, 
the  "high”  television  band,  and  the  1%-metet 
amateur  band.  The  gain  is  approximately  12 
db  over  a matched  half  wave  dipole  and  the 
beam  width  is  about  6 degrees. 

The  Feed  Line  The  recommended  feed  line 
is  an  open-wire  line  having  a 
surge  impedance  between  450  and  600  ohms. 
With  such  a line  the  VSWR  will  be  less  than 
2 to  1.  A line  with  two-inch  spacing  is  suit- 
able for  frequencies  below  100  Me.,  but  one- 
inch  spacing  (such  as  used  in  the  Gonsel  Line 
for  TV  installations)  is  recommended  for  high- 
er frequencies. 

The  Termination  If  the  array  is  to  be  used  only 
for  reception,  a suitable  ter- 
mination consists  of  two  390-ohm  carbon  re- 


6 METERS  2 METERS,  HIGH 

AND  LOW  BAND  BAND  TV,  AND 

TV  1 1/4  METERS 

S 

(side) 

90' 

32' 

L 

(length) 

166' 

10 

' 59'  4" 

W 

(Width) 

67' 

4" 

23'  11" 

1! 

wavelenths  gt 
Tilt  angle 

design  frequency 
= 680 

TABLE  I. 


488  V-H-F  and  U-H-F  Antennas 


THE  RADIO 


sistors  in  series.  If  2-watt  resistors  are  em- 
ployed, this  termination  also  is  suitable  for 
transmitter  outputs  of  10  watts  or  less.  For 
higher  powers,  however,  resistors  having  great- 
er dissipation  with  negligible  reactance  in  the 
upper  v-h-f  range  are  not  readily  available. 

For  powers  up  to  several  hundred  watts  a 
suitable  termination  consists  of  a "lossy” 
line  consisting  of  stainless  steel  wire  (corres- 
ponding to  no.  24  or  26  B&S  gauge)  spaced  2 
inches,  which  in  turn  is  terminated  by  two 
390-ohm  2-watt  carbon  resistors.  The  dissi- 
pative line  should  be  at  least  6 wavelengths 
long. 


24-8  Multi-Element  V-H-F 
Beam  Antennas 


The  rotary  multi-element  beam  is  undoubted- 
ly the  most  popular  type  of  v-h-f  antetma  in 
use.  In  general,  the  design,  assembly  and  tun- 
ing of  these  antennas  follows  a pattern  similar 
to  the  larger  types  of  rotary  beam  antennas 
used  on  the  lower  frequency  amateur  bands. 
The  characteristics  of  these  low  frequency 
beam  antennas  are  discussed  in  the  next  chap- 
ter of  this  Handbook,  and  the  information  con- 
tained in  that  chapter  applies  in  general  to  the 
v-h-f  beam  antennas  discussed  herewith. 

A Simple  Three  The  simplest  v-h-f  beam  for 
Element  Beam  the  beginner  is  the  three-ele- 
Antenna  ment  Yagi  array  illustrated  in 

figure  12.  Dimensions  are 
given  for  Yagis  cut  for  the  2-raetet  and  1/4- 
meter  bands.  The  supporting  boom  for  the  Yagi 
may  be  made  from  a smoothed  piece  of  1"  x 2 " 
wood.  The  wood  should  be  reasonably  dry  and 
should  be  painted  to  prevent  warpage  from  ex- 
posure to  sun  and  rain.  The  director  and  re- 
flector are  cut  from  lengths  of  copper  tub- 
ing, obtainable  from  any  appliance  store  that 
does  service  work  on  refrigerators.  They  should 
be  cut  to  length  as  noted  in  figure  12.  The  ele- 
ments should  then  be  given  a coat  of  aluminum 
paint.  Two  small  holes  are  drilled  at  the  center 
of  the  reflector  and  director  and  these  elements 
are  bolted  to  the  wood  boom  by  means  of  two 
1"  wood  screws.  These  screws  should  be  of 
the  plated,  or  rust-proof  variety. 

The  driven  element  is  made  of  a 78"  length 
of  % " copper  tubing,  the  ends  bent  back  upon 
each  other  to  form  a folded  dipole.  If  the  tub- 
ing is  packed  with  fine  sand  and  the  bending 
points  heated  over  a torch,  no  trouble  will  be 
had  in  the  bending  process.  If  the  tubing  does 
collapse  when  it  is  bent,  the  break  may  be  re- 
paired with  a heavy-duty  soldering  iron.  The 


FOR  METERS 

0 = ZZ" 

D-A  = 9' 

A-23^*' 

A-R-9' 

R=25* 

bend  1 • 

RADIUS^ 

R“  REFLECTOR 
40" LONG 


OAINTSDR  (20- LONG  FOR  meters) 


FLATTEN  ENDS  OF 
TUBING  AND  DRILL 
FOR  6/32  SCREWS 


Figure  12 

SIMPLE  3.ELEMENT  BEAM  FOR  2 AND 
1'4  METERS 


driven  element  is  next  attached  to  the  center 
of  the  wood  boom,  mounted  atop  a small  in- 
sulating plate  made  of  bakelite,  micarta  or 
some  other  non-conducting  material.  It  is  held 
in  place  in  the  same  manner  as  the  parasitic 
elements.  The  two  free  ends  of  the  folded  di- 
pole are  hammered  flat  and  drilled  for  a 6-32 
bolt.  These  bolts  pass  through  both  the  insu- 
lating block  and  the  boom,  and  hold  the  free 
tips  of  the  element  in  place. 

A length  of  75-ohra  Twin-Lead  TV-type  line 
should  be  used  with  this  beam  antenna.  It  is 
connected  to  each  of  the  free  ends  of  the  folded 
dipole.  If  the  .antenna  is  mounted  in  the  verti- 
cal plane,  the  75-ohm  line  should  be  brought 
away  from  the  antenna  for  a distance  of  four 
to  six  feet  before  it  drops  down  the  tower  to 
lessen  interaction  between  the  antenna  ele- 
ments and  the  feed  line.  The  complete  antenna 
is  light  enough  to  be  turned  by  a TV  rotator. 

A simple  Yagi  antenna  of  this  type  will  pro- 
vide a gain  of  7 db  over  the  entire  2-meter  or 
l!4-meter  band,  and  is  highly  recommended  as 
an  "easy-to-build”  beam  for  the  no  vice  or 
beginner. 

An  S-Element  Figures  13  and  14  illus- 

"Tippoble"  Arroy  trate  an  8-element  rotary 

for  144  Me.  array  for  use  on  the  144- 

Mc.  amateur  band.  This 
array  is  "tippable”  to  obtain  either  horizontal 
or  vertical  polarization.  It  is  necessary  that 
the  transmitting  and  receiving  station  use  the 
same  polarization  for  the  ground-wave  signal 
propagation  which  is  characteristic  of  this  fre- 


HANDBOOK 


VHF  Parasitic  Arrays 


489 


FtLE  END  TO  FIT 


MAIN  BOOMS  - -k  APPROX.  O.D. 


RG-59/U  CABLES  iqflclOFFSET PIVOT \ 
EACH  40- LONG  HEAD.  1/4  " I 

n*^^B3oural  1 


RG-e/U  CABLE-^' 
TO  '“T-  COAXIAL 
FITTING- 


INSULATING  ROD.  ENDS 
CUT  DOWN  TO  GO  INTO  TUBING 
ABOUT  -5"/ 


ENOS  OF  TUBING 
WOOD  DOWELS  IN- 
SIDE FOR  STRENGTH 


AS  SHOWN,  ANTENNA  IS 
HORIZONTALLY  POLARIZED. 


PULL  TO  SWING  MAIN  BOOM  90“ 
FOR  VERTICAL  POLARITY, 


RADIAL  BEARING 


Figure  13 

CONSTRUCTIONAL  DRAWING  OF  AN  EIGHT-ELEMENT  TIPPABLE  144.MC.  ARRAY 


quency  range.  Although  polarization  has  been 
loosely  standardized  in  various  areas  of  the 
country,  exceptions  are  frequent  enough  so 
that  it  is  desirable  that  the  polarization  of  an- 
tenna radiation  be  easily  changeable  from  hori- 
zontal to  vertical. 

The  antenna  illustrated  has  shown  a signal 
gain  of  about  11  db,  representing  a power  gain 
of  about  13.  Although  the  signal  gain  of  the 


antenna  is  the  same  whether  it  is  oriented  for 
vertical  or  horizontal  polarization,  the  hori- 
zontal beam  width  is  smaller  when  the  antenna 
is  oriented  for  vertical  polarization.  Conver- 
sely, the  vertical  pattern  is  the  sharper  when 
the  antenna  system  is  oriented  for  horizontal 
polarization. 

The  changeover  from  one  polarization  to  the 
other  is  accomplished  simply  by  pulling  on  the 


490  V-H-F  and  U-H-F  Antennas 


THE  RADIO 


appropriate  cord.  Hence,  the  operation  is  based  13.  However,  the  pointers  given  in  the  follow- 

on  the  offset  head  sketched  in  figure  13.  Al-  ing  paragraphs  will  be  of  assistance  to  those 

though  a wood  mast  has  been  used,  the  same  wishing  to  reproduce  the  array, 
system  may  be  used  with  a pipe  mast.  The  drilling  of  holes  for  the  small  elements 

The  40-inch  lengths  of  RG-59/U  cable  (elec-  should  be  done  carefully  on  accurately  marked 

trically  ^ wavelength)  running  from  the  center  centers.  A small  angular  error  in  the  drilling 

of  each  folded  dipole  driven  element  to  the  of  these  holes  will  result  in  a considerable 

coaxial  T-junction  allow  enough  slack  to  per-  misalignment  of  the  elements  after  the  array  is 

mit  free  movement  of  the  main  boom  when  assembled.  The  same  consideration  is  true  of 

changing  polarity.  Type  RG-8/U  cable  is  run  the  filing  out  of  the  rounded  notches  in  the 

from  the  T-junction  to  the  operating  position.  ends  of  the  main  boom  for  the  fitting  of  the 

Measured  standing-wave  ratio  was  less  than  two  antenna  booms. 

2:1  over  the  144  to  148  Me.  band,  with  the  Short  lengths  of  wood  dowel  are  used  freely 

lengths  and  spacings  given  in  figure  13.  in  the  construction  of  the  array.  The  ends  of 

the  small  elements  are  plugged  with  an  inch 
Construction  of  Most  of  the  constructional  or  so  of  dowel,  and  the  ends  of  the  antenna 

the  Array  aspects  of  the  antenna  array  booms  are  similarly  treated  with  larger  discs 

are  self-evident  from  figure  pressed  into  place. 


HANDBOOK 


VHF  Parasitic  Arrays  491 


The  ends  of  the  folded  dipoles  are  made  in 
the  following  manner:  Drive  a length  of  dowel 
into  the  short  connecting  lengths  of  aluminum 
tubing.  Then  drill  down  the  center  of  the  dowel 
with  a clearance  hole  for  the  connecting  screw. 
Then  shape  the  ends  of  the  connecting  pieces 
to  fit  the  sides  of  the  element  ends.  After  as- 
sembly the  junctions  may  be  dressed  with  a 
file  and  sandpaper  until  a smooth  fit  is  ob- 
tained. 

The  mast  used  for  supporting  the  array  is  a 
30-foot  spliced  2 by  2.  A large  discarded  ball 
bearing  is  used  as  the  radial  load  bearing  and 
guy-wire  termination.  Enough  of  the  upper-mast 
corners  were  removed  with  a draw-knife  to  per- 
mit sliding  the  ball  beating  down  about  9 feet 
from  the  top  of  the  mast.  The  bearing  then  was 
encircled  by  an  assembly  of  three  pieces  of 
dural  ribbon  to  form  a clamp,  with  ears  for 
tightening  screws  and  attachment  of  the  guy 
wires.  The  bearing  then  was  greased  and  cov- 
eted with  a piece  of  auto  inner  tube  to  serve 
as  protection  from  the  weather.  Another  junk- 
box  bearing  was  used  at  the  bottom  of  the  mast 
as  a thrust  bearing. 

The  main  booms  were  made  from  J^-inch 
aluminum  electrical  conduit.  Any  size  of  small 
tubing  will  serve  for  making  the  elements. 
Note  that  the  main  boom  is  mounted  at  the  bal- 
ance center  and  not  necessarily  at  the  physi- 
cal center.  The  pivot  bolt  in  the  offset  head 
should  be  tightened  sufficiently  that  there  will 
be  adequate  friction  to  hold  the  array  in  posi- 
tion. Then  an  additional  nut  should  be  placed 
on  the  pivot  bolt  as  a lock. 

In  connecting  the  phasing  sections  between 
the  T-junction  and  the  centers  of  the  folded 
dipoles,  it  is  important  that  the  center  con- 
ductors of  the  phasing  sections  be  connected 
to  the  same  side  of  the  driven  elements  of  the 
antennas.  In  other  words,  when  the  antenna  is 
oriented  for  horizontal  polarization  and  the 
center  of  the  coaxial  phasing  section  goes  to 
the  left  side  of  the  top  antenna,  the  center 
conductor  of  the  other  coaxial  phasing  section 
should  go  to  the  left  side  of  the  bottom  an- 
tenna. 

The  “Screen  Beam**  This  highly  effective  ro- 
for  2 Meters  tary  array  for  the  144  Me. 

amateur  band  was  de- 
signed by  the  staff  of  the  Experimental  Phy- 
sics Laboratory,  The  Hague,  Netherlands  for 
use  at  the  2 meter  experimental  station  PEIPL. 
The  array  consists  of  10  half  wave  radiators 
fed  in  phase,  and  arranged  in  two  stacked  tows 
of  five  radiators.  0.2  wavelength  behind  this 
plane  of  radiators  is  a reflector  screen,  meas- 
uring approximately  15'  x 9'  in  size.  The  an- 
tenna provides  a power  gain  of  15  db,  and  a 
front  to  back  ratio  of  approximately  28  db. 


Figure  15 

DETAIL  OF  LAYOUT  AND  DIMENSIONS 
OF  BEAM  ASSEMBLY  OF  PEIPL 


The  10  dipoles  are  fed  in  phase  by  means 
of  a length  of  balanced  transmission  line,  a 
quartet-wave  matching  trtmsformer,  and  a ba- 
lun.  A 72-ohm  coaxial  line  couples  the  array 
to  the  transmitter.  A drawing  of  the  array  is 
shown  in  figure  15. 

The  reflecting  screen  measures  14'  9"  high 
by  8'  4"  wide,  and  is  made  of  welded  " dia- 
meter steel  tubing.  Three  steel  reinforcing 
bars  are  welded  horizontally  across  the  frame- 
work directly  behind  each  pair  of  horizontal 
dipoles.  The  intervening  spaces  ate  filled 
with  lengths  of  no.  12  enamel-coated  copper 
wire  to  complete  the  screen.  The  spacing  be- 
tween the  wires  is  2".  Four  cross  braces  ate 
welded  to  the  corners  of  the  frame  for  addi- 
tional bracing,  and  a single  vertical  % " rod 
runs  up  the  middle  of  the  frame.  The  complete, 
welded  frame  is  shown  in  figure  15.  The  no. 
12  screening  wires  are  tun  between  6-32  bolts 
placed  in  holes  drilled  in  each  outside  verti- 
cal member  of  the  frame. 

The  antenna  assembly  is  supported  away 
from  the  reflector  screen  by  means  of  ten 
lengths  of  steel  tubing,  each  1'  3/4"  long. 


492  V-H-F  and  U-H-F  Antennas 


THE  RADIO 


WOOD  BLOCK 


Figure  16 

THE  MOUNTING  BLOCK  FOR  EACH  SET 
OF  ELEMENTS 


These  tubes  are  welded  onto  the  center  tube 
of  each  group  of  three  horizontal  bracing  tubes, 
and  are  so  located  to  support  the  horizontal  di- 
pole at  its  exact  center.  The  dipoles  are  at- 
tached to  the  supporting  rods  by  means  of 
small  phenolic  insulating  blocks,  as  shown  in 
figure  16.  The  radiators  are  therefore  insulated 
from  the  screen  reflector.  The  inner  tips  of 
the  radiators  are  held  by  small  polystyrene 
blocks  for  rigidity,  and  ate  cross  connected  to 
each  other  by  a transposed  length  of  TV-type 
400  ohm  open  wire  line.  The  entire  array  is 
fed  at  the  point  A-A,  illustrated  in  figure  15. 

The  matching  system  for  the  beam  is  mounted 
behind  the  reflector  screen , and  is  shown  in 
figure  17.  A quarter- wave  transformer  (B)  drops 
the  relatively  high  impe.dance  of  the  antenna 
array  to  a suitable  value  for  the  low  imped- 
ance balun  (D).  An  adjustable  matching  stub 
(C)  and  two  variable  capacitors  (C,  and  C,) 
ate  employed  for  impedance  matching.  The 
two  variable  capacitors  are  mounted  in  a 


Figure  18 

HORIZONTAL  RADIATION  PATTERN  OF 
THE  PEIPL  ARRAY.  THE  FRONT-TO- 
BACK  RATIO  IS  ABOUT  28  db  IN  AMPLI- 
TUDE, AND  THE  FORWARD  GAIN  AP- 
PROXIMATELY 15  db. 


Figure  17 

THE  MATCHING  UNIT  IN  DETAIL  FOR 
THE  PEIPL  BEAM  DESIGN,  WHICH  AL- 
LOWS THE  USE  OF  72.0HM  COAX 


watertight  box,  with  the  balun  and  matching 
stubs  entering  the  bottom  and  top  of  the  box, 
respectively. 

The  matching  procedure  is  carried  out  by 
the  use  of  a standing  wave  meter  (SWR  bridge). 
A few  watts  of  power  are  f e d to  the  array 
through  the  SWR  meter,  and  the  setting  of  the 
shorting  stub  on  C and  the  setting  of  the  two 
variable  capacitors  are  adjusted  for  lowest 
SWR  at  the  chosen  operating  frequency.  The 
capacity  settings  of  the  two  variable  capaci- 
tors should  be  equal.  The  final  adjustment  is 
to  set  the  shorting  stub  of  the  balun  (D)  to  re- 
move any  residual  reactance  that  might  appear 
on  the  transmission  line.  With  proper  adjust- 
ment, the  VSWR  of  the  array  may  be  held  to 
less  than  1.5  to  1 over  a 2 megacycle  range 
of  the  2-meter  band. 

The  horizontal  radiation  pattern  of  this  array 
is  shown  in  figure  18. 

Long  Yogi  For  a given  power  gain,  the 

Antennas  Vflgf  antenna  can  be  built 

lighter,  more  compact,  and 
with  less  wind  resistance  than  any  other  type. 


HANDBOOK 


VHF  Parasitic  Arrays  493 


ELEMENT  DIMENSIONS.  2 METER  BAND 


ELEMENT 

(D/AM.  1^8') 

LENGTH 

SPACING 

FROM 

Dl  POLE 

I44MC. 

145  MC. 

148  MC. 

147  MC. 

REFLECTOR 

41" 

-r 

19' 

DIRECTORS 

36 

36^' 

36  r 

36^' 

Dl=  7" 

DRIVEN  ELEMENT 

ae.s* 


-#6  WIRE  FOR  300  (L 
MATCH. 

C10WIRE  FOR  450 
MATCH, 


21 


INSULATING  ^FLATTEN 
PLATE  TUBING 

AT  ENDS. 


D2=  14.5' 
D3S  22" 
D4=  38" 
05=  70" 

D6=  102' 

07*  134' 

De*  i6«" 
09* 198" 
Di0*230" 
Di1  = 242" 


Figure  19 

DESIGN  DIMENSIONS  FOR  A 2-METER  “LONG  YAGI"  ANTENNA 


On  the  other  hand,  if  a Yagi  array  of  the  same 
approximate  size  and  weight  as  another  an- 
tenna type  is  built,  it  will  provide  a higher 
order  of  power  gain  and  directivity  than  that 
of  the  other  antenna. 

The  p)ower  gain  of  a Yagi  antenna  increases 
directly  with  the  physical  length  of  the  array. 
The  maximum  practical  length  is  entirely  a 
mechanical  problem  of  physically  supporting 
the  long  series  of  director  elements,  although 
when  the  array  exceeds  a few  wavelengths  in 
length  the  element  lengths,  spacings,  and 
Q*s  become  more  and  more  critical.  The  ef- 
fectiveness of  the  array  depends  upon  a proper 
combination  of  the  mutual  coupling  loops 
between  adjacent  directors  and  between  the 
first  director  and  the  driven  element. 

Practically  all  work  on  Yagi  antennas  with 
more  than  three  or  four  elements  has  been  on 
an  experimental,  cut-and-try  basis.  Figure  19 


provides  dimensions  for  a typical  Long  Yagi 
antenna  for  the  2-meier  VHF  band.  Note  that 
all  directors  have  the  same  physical  length. 
If  the  long  Yagi  is  designed  so  that  the  di- 
rectors gradually  decrease  in  length  as  they 
progress  from  the  dipole  bandwidth  will  be 
increased,  and  both  side  lobes  and  forward 
gain  will  be  reduced.  One  advantage  gained 
from  staggered  director  length  is  that  the 
array  can  be  shortened  and  lengthened  by 
adding  or  taking  away  directors  without  the 
need  for  retuning  the  remaining  group  of  para- 
sitic elements.  When  all  directors  are  the 
same  length,  they  must  be  all  shortened  en 
masse  as  the  array  is  lengthened,  and  vice- 
versa  when  the  array  is  shortened. 

A full  discussion  of  Long  Yagi  antennas, . 
including  complete  design  and  construction 
information  may  be  had  in  the  VHF  Handbook, 
available  through  Radio  Publications,  Inc., 
Wilton.,  Conn. 


CHAPTER  TWENTY-FIVE 


Rotary  Beams 


The  rotatable  anteraia  array  has  become  al- 
most standard  equipment  for  operation  on  the 
28-Mc.  and  50-Mc.  bands  and  is  commonly  used 
on  the  14-Mc.  and  21-Mc.  bands  and  on  those 
frequencies  above  144  Me.  The  rotatable  array 
offers  many  advantages  for  both  military  and 
amateur  use.  The  directivity  of  the  antenna 
types  commonly  employed,  particularly  the 
unidirectional  arrays,  offers  a worthwhile  re- 
duction in  interference  from  undesired  direc- 
tions. Also,  the  increase  in  the  ratio  of  low- 
angle  radiation  plus  the  theoretical  gain  of 
such  arrays  results  in  a relatively  large  in- 
crease in  both  the  transmitted  signal  and  the 
signal  intensity  from  a station  being  received. 

A significant  advantage  of  a rotatable  an- 
tenna array  in  the  case  of  the  normal  sration  is 
that  a relatively  small  amount  of  space  is  re- 
quired for  erection  of  the  antenna  system.  In 
fact,  one  of  the  best  types  of  installation  uses 
a single  telephone  pole  with  the  rotating  struc- 
ture holding  the  antenna  mounted  atop  the  pole. 
To  obtain  results  in  all  azimuth  directions 
from  fixed  arrays  comparable  to  the  gain  and 
directivity  of  a single  rotatable  three-element 
parasitic  beam  would  requite  several  acres  of 
surface. 

There  ate  two  normal  configurations  of  radi- 
ating elements  which,  when  horizontally  polar- 
ized, will  contribute  to  obtaining  a low  angle 
of  radiation.  These  configurations  are  the  end- 
fire  array  and  the  broadside  array.  Tbe  con- 


ventional three-  or  four-element  rotary  beam 
may  properly  be  called  a unidirectional  para- 
sitic end-fire  array,  and  is  actually  a type  of 
yagi  array.  The  flat-top  beam  is  a type  of  bi- 
directional end-fire  array.  The  broadside  type 
of  array  is  also  quite  effective  in  obtaining 
low-angle  radiation,  and  although  widely  used 
in  FM  and  TV  broadcasting  has  seen  little  use 
by  amateur  stations  in  rotatable  arrays. 


25-1  Unidirectional 

Parasitic  End-Fire  Arrays 

(Yogi  Type) 

If  a single  parasitic  element  is  placed  on 
one  side  of  a driven  dipole  at  a distance  of 
from  0.1  to  0.25  wavelength  the  parasitic  ele- 
ment can  be  tuned  to  make  the  array  substan- 
tially unidirectional. 

This  simple  array  is  termed  a two  element 
parasitic  beam. 


25-2  The  Two  Element  Beam 

The  two  element  parasitic  beam  provides 
the  greatest  amount  of  gain  per  unit  size  of 
any  array  commonly  used  by  radio  amateurs. 


4 94 


Parasitic  Arrays  4'95 


Figure  1 

GAIN  VS  ELEMENT  SPACING  FOR  A TWO- 
ELEMENT  CLOSE-SPACED  PARASITIC 
BEAM  ANTENNA  WITH  PARASITIC  ELE- 
MENT OPERATING  AS  A DIRECTOR  OR 
REFLECTOR 


Figure  2 

RADIATION  RESISTANCE  AS  A FUNCTION 
OF  ELEMENT  SPACING  FOR  A TWO-ELE- 
MENT PARASITIC  ARRAY 


Such  an  antenna  is  capable  of  a signal  gain 
of  5 db  ovet  a dipole,  with  a front-to-back  ratio 
of  7 db  to  15  db,  depending  upon  the  adjust- 
ment of  the  parasitic  element.  The  parasitic 
element  may  be  used  either  as  a director  or 
as  a reflector. 

The  optimum  spacing  for  a reflector  in  a 
two-element  array  is  approximately  0.13  wave- 
length and  with  optimum  adjustment  of  the 
length  of  the  reflector  a gain  of  approximately 
5 db  will  be  obtained,  with  a feed-point  resist- 
ance of  about  25  ohms. 

If  the  parasitic  element  is  to  be  used  as  a 
director  the  optimum  spacing  between  it  and 
the  driven  element  is  0.11  wavelength.  The 
gain  will  theoretically  be  slightly  greater  than 
with  the  optimum  adjustment  for  a reflector 
(about  5.5  db)  and  the  radiation  resistance 
will  be  in  the  vicinity  of  17  ohms. 

The  general  characteristics  of  a two-element 
parasitic  array  may  be  seen  in  figures  1,  2 and 
3.  The  gain  characteristics  of  a two-element 
array  when  the  parasitic  element  is  used  as  a 
director  or  as  a reflector  are  shown.  It  can  be 
seen  that  the  director  provides  a maximum  of 
5.3  db  gain  at  a spacing  of  slightly  greater 
than  0.1  wavelength  from  the  antenna.  In  the 
interests  of  greatest  power  gain  and  size  con- 
servation, therefore,  the  choice  of  a parasitic 
director  would  be  wiser  than  the  choice  of  a 
parasitic  reflector,  although  the  gain  differ- 
ence between  the  two  is  small. 

Figure  2 shows  the  relationship  between 
the  element  spacing  and  the  radiation  resist- 


ance for  the  two  element  parasitic  array  for 
both  the  reflector  and  the  director  case.  Since 
the  optimum  antenna-director  spacing  for  maxi- 
mum gain  results  in  an  antenna  radiation  re- 
sistance of  about  17  ohms,  and  the  optimum 
antenna-reflector  spacing  for  maximum  gain 
results  in  an  antenna  radiation  resistance  of 
about  25  ohms,  it  may  be  of  advantage  in  some 
instances  to  choose  the  antenna  with  the  high- 
er radiation  resistance,  assuming  other  fac- 
tors to  be  equal. 

Figure  3 shows  the  front-to-back  ratio  for 
the  two  element  parasitic  array  for  both  the 
reflector  and  director  cases.  To  produce  these 
curves,  the  elements  were  tuned  for  maximum 
gain  of  the  array.  Better  front-to-back  ratios 
may  be  obtained  at  the  expense  of  array  gain, 
if  desired,  but  the  general  shape  of  the  curves 
remains  the  same.  It  can  be  readily  observed 
that  operation  of  the  parasitic  element  as  a 
reflector  produces  relatively  poor  front-to- 
back  ratios  except  when  the  element  spacing 
is  greater  than  0.15  wavelength.  However,  at 
this  element  spacing,  the  gain  of  the  array  be- 
gins to  suffer. 

Since  a radiation  resistance  of  17  ohms  is 
not  unduly  hard  to  match,  it  can  be  argued  that 
the  best  all-around  performance  may  be  ob- 
tained from  a two  element  parasitic  beam  em- 
ploying 0.11  element  spacing,  with  the  para- 
sitic element  tuned  to  operate  as  a director. 
This  antenna  will  provide  a forward  gain  of 
5.3  db,  with  a front-to-back  ratio  of  10  db,  or 
slightly  greater.  Closer  spacing  than  0.11 


4 96  Rotary  Beams 


THE  RADIO 


wavelength  may  be  employed  for  greater  front- 
to-back  ratios,  but  the  radiation  resistance  of 
the  array  becomes  quite  low,  the  bandwidth 
of  the  array  becomes  very  narrow,  and  the  tun- 
ing becomes  quite  critical.  Thus  the  Q of  the 
antenna  system  will  be  increased  as  the  spac- 
ing between  the  elements  is  decreased,  and 
smaller  optimum  frequency  coverage  will 
result. 


Element  Lengths  When  the  parasitic  element 
of  a two-element  array  is 
used  as  a director,  the  following  formulas  may 
be  used  to  determine  the  lengths  of  the  driven 
element  and  the  parasitic  director,  assuming 
an  element  diameter-to-length  ratio  of  200  to 
400: 


476 

Driven  element  length  (feet)  = 

FMc. 

450 

Director  length  (feet)  = 

FMc. 


Element  spacing  (feet)  = 


Figure  4 

FIVE  ELEMENT  28  MC  BEAM 
ANTENNA  AT  W6SAI 


Antenna  boom  Is  made  of  twenty  foot 
length  of  Sears,  Roebuck  Co.  three- 
inch  aluminum  Irrigation  pipe.  Spacing 
between  elements  is  five  feet.  Ele- 
ments are  made  of  twelve  foot  lengths 
of  7/8-ineh  aluminum  tubing,  with  ex- 
tension tips  made  of  3/4-inch  tubing. 
Gamma  matching  device,  element 
clamps,  and  "Oxen  Yoke"  el ement-to- 
boom  clamps  are  made  by  Continental 
Electron! cs  & Sound  Co.,  Doyfon  27, 
Ohio,  Beam  dimensions  are  taken  from 


Figure  3 

FRONT-TO-BACK  RATIO  AS  A FUNCTION 
OF  ELEMENT  SPACING  FOR  A TWO-ELE- 
MENT PARASITIC  ARRAY 


The  effective  bandwidth  taken  between  the 
1.5/1  standing  wave  points  of  an  array  cut  to 
the  above  dimensions  is  about  2.5%  of  the 
operating  frequency.  This  means  that  an  array 
pre-cut  to  a frequency  of  14,150  kilocycles 
would  have  a bandwidth  of  350  kilocycles  (plus 
or  minus  175  kilocycles  of  the  center  frequen- 
cy), and  therefore  would  be  effective  over  the 
whole  20  meter  band.  In  like  fashion,  a 15 
meter  array  should  be  pre-cut  to  21,200  kilo- 
cycles. 

A beam  designed  for  use  on  the  10-metet 
band  would  have  an  effective  bandwidth  of 
some  700  kilocycles.  Since  the  10-meter  band 
is  1700  kilocycles  in  width,  the  array  should 
either  be  cut  to  28,500  kilocycles  for  opera- 
tion in  the  low  frequency  portion  of  the  band, 
or  to  29,200  kilocycles  for  operation  in  the 
high  frequency  portion  of  the  band.  Operation 
of  the  antenna  outside  the  effective  bandwidth 
will  increase  the  SWR  on  the  transmission 
line,  and  noticeably  degrade  both  the  gain  and 
front-to-back  ratio  performance.  The  height 
above  ground  also  influences  the  F/B  ratio. 


25-3  The  Three-Element  Array 


The  three-element  array  using  a director, 
driven  element,  and  reflector  will  exhibit  as 
much  as  30  db  front-to-back  ratio  and  20  db 
front-to-side  ratio  for  low-angle  radiation.  The 
theoretical  gain  is  about  9 db  over  a dipole  in 
free  space.  In  actual  practice,  the  array  will 
often  show  7 to  10  db  apparent  gain  over  a 
horizontal  dipole  placed  the  same  height  above 
ground  (at  28  and  14  Me.). 

The  use  of  more  than  three  elements  is  de- 
sirable when  the  length  of  the  supporting  struc- 
ture is  such  that  spacings  of  approximately 


HANDBOOK 


Parasitic  Arrays  4 97 


0.2  wavelength  between  elements  becomes 
possible.  Four-element  arrays  are  quite  com- 
mon on  the  28-Mc.  and  50-Mc.  bands,  and  five 
elements  are  sometimes  used  for  increased 
gain  and  discrimination.  As  the  number  of  ele- 
ments is  increased  the  gain  and  front-to-back 
ratio  increases  but  the  radiation  resistance  de- 
creases and  the  bandwidth  or  frequency  range 
over  which  the  antenna  will  operate  without 
reduction  in  effectiveness  is  decreased. 


Material  for  While  the  elements  may  consist 
Elements  of  wire  supported  on  a wood 
framework,  self-supporting  ele- 
ments of  tubing  are  much  to  be  preferred.  The 
latter  type  array  is  easier  to  construct,  looks 
better,  is  no  more  expensive,  and  avoids  the 
problem  of  getting  sufficiently  good  insulation 
at  the  ends  of  the  elements.  The  voltages 
reach  such  high  values  towards  the  ends  of 
the  elements  that  losses  will  be  excessive, 
unless  the  insulation  is  excellent. 

The  elements  may  be  fabricated  of  thin- 
walled  steel  conduit,  or  hard  drawn  thin-walled 
copper  tubing,  but  dural  tubing  is  much  better. 
Or,  if  you  prefer,  you  may  purchase  tapered 
copper-plated  steel  tubing  elements  designed 
especially  for  the  purpose.  Kits  are  available 
complete  with  rotating  mechanism  and  direction 
indicator,  for  those  who  desire  to  purchase 
the  whole  system  ready  to  put  up. 


Element  Spacing  The  optimum  spacing  for  a 
two-element  array  is,  as  has 
been  mentioned  before,  approximately  0.11 
wavelength  for  a director  and  0.13  wavelength 
for  a reflector.  However,  when  both  a director 
and  a reflector  are  combined  with  the  driven 
element  to  make  up  a three-element  array  the 
optimum  spacing  is  established  by  the  band- 
width which  the  antenna  will  be  required  to 
cover.  Wide  spacing  (of  the  order  of  0.25  wave- 
length between  elements)  will  result  in  greater 
bandwidth  for  a specified  maximum  standing- 
wave  ratio  on  the  antenna  transmission  line. 
Smaller  spacings  may  be  used  when  boom 
length  is  an  important  consideration,  but  for  a 
specified  standing-wave  ratio  and  forward  gain 
the  frequency  coverage  will  be  smaller.  Thus 
the  Q of  the  antenna  system  will  be  increased 
as  the  spacing  between  the  elements  is  de- 
creased, resulting  in  smaller  frequency  cover- 
age, and  at  the  same  time  the  feed-point  im- 
pedance of  the  driven  element  will  be  de- 
creased. 

For  broad-band  coverage,  such  as  the  range 
from  26.96  to  29-7  Me.  or  from  50  to  54  Me., 
0.2  wavelength  spacing  from  the  driven  ele- 
ment to  each  of  the  parasitic  elements  is  rec- 


ommended. For  narrower  bandwidth,  such  as 
would  be  adequate  for  the  14.0  to  14.4  Me. 
band  or  the  144  to  148  Me.  band,  the  radiator 
to  parasitic  element  spacing  may  be  reduced 
to  0.12  wavelength,  while  still  maintaining 
adequate  array  bandwidth  for  the  amateur  band 
in  question. 

Length  of  the  Experience  has  shown  that 

Parasitic  Elements  it  is  practical  to  cut  the 
prarsitic  elements  of  a 
three-elementparasitic  array  to  a predetermined 
length  before  the  installation  of  such  an  an- 
tenna. A pte-tuned  antenna  such  as  this  will 
give  good  signal  gain,  adequate  front-to-back 
ratio,  and  good  bandwidth  factor.  By  carefully 
tuning  the  array  after  it  is  in  position  the  gain 
may  be  increased  by  a fraction  of  a db,  and 
the  front-to-back  ratio  by  several  db.  However 
the  slight  improvement  in  performance  is  us- 
ually not  worth  the  effort  expended  in  tuning 
time. 

The  closer  the  lengths  of  the  parasitic  ele- 
ments are  to  the  resonant  length  of  the  driven 
element,  the  lower  will  be  the  feed-point  resist- 
ance of  the  driven  element,  and  the  smaller 
will  be  the  bandwidth  of  the  array.  Hence,  for 
wide  frequency  coverage  the  director  should 
be  considerably  shorter,  and  the  reflector  con- 
siderably longer  than  the  driven  element.  For 
example,  the  director  should  still  be  less  than 
a resonant  half  wave  at  the  upper  frequency 
limit  of  the  range  wherein  the  antenna  is  to  be 
operated,  and  the  reflector  should  still  be  long 
enough  to  act  as  a reflector  at  the  lower  fre- 
quency limit.  Another  way  of  stating  the  same 
thing  is  to  say,  in  the  case  of  an  array  to  covet 
a wide  frequency  range  such  as  the  amateur 
range  from  26.96  to  29.7  Me.  or  the  width  of  a 
low-band  TV  channel,  that  the  director  should 
be  cut  for  the  upper  end  of  the  band  and 
the  reflector  for  the  lower  end  of  the  band.  In 
the  case  of  the  26.96  to  29.7  Me.  range  this 
means  that  the  director  should  be  about  8 pet 
cent  shorter  than  the  dfiven  element  and  the 
reflector  should  be  about  8 pet  cent  longer. 
Such  an  antenna  will  show  a relatively  con- 
stant gain  of  about  6 db  over  its  range  of  cov- 
erage, and  the  pattern  will  not  reverse  at  any 
point  in  the  range. 

Where  the  frequency  range  to  be  covered  is 
somewhat  less,  such  as  a high-band  TV  chan- 
nel, the  14.0  to  14.4  Me.  amateur  band,  or  the 
lower  half  of  the  amateur  28-Mc.  phone  band, 
the  reflector  should  be  about  5 per  cent  longer 
than  the  driven  element,  and  the  director  about 
5 pet  cent  shorter.  Such  an  antenna  will  per- 
form well  over  its  rated  frequency  band,  will 
not  reverse  its  pattern  over  this  band,  and  will 
show  a signal  gain  of  7 to  8 db.  See  figure  5 
for  design  figures  for  3-element  arrays. 


498  Rotary  Beams 


THE  RADIO 


TYPE 

PRIVEN  ELEMENT 

REFLECTOR 

iST  DIRECTOR 

2N0  DIRECTOR 

3RD  DIRECTOR 

SPACING  BET- 

APPR0)LCAIN 

APPROX.  RADIATION 

LENGTH 

LENGTH 

LENGTH 

LENGTH 

LENGTH 

WEEN  ELEMENTS 

DB 

RESISTANCE  (A) 

3-element 

473 

501 

445 

.1  5 - .1  5 

7.5 

20 

F (MC) 

F (.MC) 

f (uc) 

B-element 

F (»^) 

.^01. 

F(MC) 

F (MC) 

— 

— 

.25-. 25 

e.5 

35 

4-element 

■473_ 

501 

■*50 

450 

.2-.2-.2 

9.5 

20 

F(MC) 

F(mc) 

F(uc) 

F(MC) 

6-element 

473 

501 

450 

450 

4^.0 

.2-  .2 -.2-2 

10.0 

1 5 

F(MC) 

F(mc) 

F(mc) 

F(MC) 

F(MC) 

Figure  5 

DESIGN  CHART  FOR  PARASITIC  ARRAYS  (DIMENSIONS  GIVEN  IN  FEET) 


More  Than  A small  amount  of  additional 

Three  Elements  gain  may  be  obtained  through 
use  of  more  than  two  parasitic 
elements,  at  the  ettpense  of  reduced  feed-point 
impedance  and  lessened  bandwidth.  One  addi- 
tional director  will  add  about  1 db,  and  a sec- 
ond additional  director  (making  a total  of  five 
elements  including  the  driven  element)  will 
add  slightly  less  than  one  db  more.  In  the 
v-h-f  range,  where  the  additional  elements  may 
be  added  without  much  difficulty,  and  where 
required  bandwidths  are  small,  the  use  of  mote 
than  two  parasitic  elements  is  quite  practic- 
able. 

Stacking  of  Parasitic  arrays  (yagis)  may 
Yogi  Arrays  be  stacked  to  provide  addition- 
al gain  in  the  same  manner  that 
dipoles  may  be  stacked.  Thus  if  an  array  of 
six  dipoles  would  give  a gain  of  10  db,  the 
substitution  of  yagi  arrays  for  each  of  the  di- 
poles would  add  the  gain  of  one  yagi  array  to 
the  gain  obtained  with  the  dipoles.  However, 
the  yagi  arrays  must  he  more  widely  spaced 
than  the  dipoles  to  obtain  this  theoretical  im- 
provement. As  an  example,  if  six  5-element 
yagi  arrays  having  a gain  of  about  10  db  were 
substituted  for  the  dipoles,  with  appropriate 
increase  in  the  spacing  between  the  arrays, 
the  gain  of  the  whole  system  would  approach 
the  sum  of  the  two  gains,  or  20  db.  A group  of 
arrays  of  yagi  antennas,  with  recommended 
spacing  and  approximate  gains,  ate  illus- 
trated in  figure  6. 


25-4  Feed  S)rstems  for 

Parasitic  (Yogi)  Arrays 

The  table  of  figure  5 gives,  in  addition  to 
other  information,  the  approximate  radiation 
resistance  referred  to  the  center  of  the  driven 
element  of  multi-element  parasitic  arrays.  It  is 
obvious,  from  these  low  values  of  radiation 


resistance,  that  especial  care  must  be  taken 
in  materials  used  and  in  the  construction  of 
the  elements  of  the  array  to  insure  that  ohmic 
losses  in  the  conductors  will  not  be  an  appre- 
ciable percentage  of  the  radiation  resistance. 
It  is  also  obvious  that  some  method  of  imped- 
ance transformation  must  be  used  in  many 
cases  to  match  the  low  radiation  resistance 
of  these  antenna  arrays  to  the  normal  range  of 
characteristic  impedance  used  for  antenna 
transmission  lines. 

A group  of  possible  methods  of  impedance 
matching  is  shown  in  figures  7,  8,  9 and  10. 
All  these  methods  have  been  used  but  certain 
of  them  offer  advantages  over  some  of  the 
other  methods.  Generally  speaking  it  is  not 
mechanically  desirable  to  break  the  center  of 
the  driven  element  of  an  array  for  feeding  the 
system.  Breaking  the  driven  element  rules  out 
the  practicability  of  building  an  all-metal  or 
"plumber’s  delight’’  type  of  array,  and  im- 
poses mechanical  limitations  with  any  type  of 
construction.  However,  when  continuous  rota- 
tion is  desired,  an  arrangement  such  as  shown 
in  figure  9D  utilizing  a broken  driven  element 
with  a rotatable  transformer  for  coupling  from 
the  antenna  transmission  line  to  the  driven 
element  has  proven  to  be  quite  satisfactory. 
In  fact  the  method  shown  in  figure  9D  is  prob- 
ably the  most  practicable  method  of  feeding 
the  driven  element  when  continuous  rotation 
of  die  antenna  array  is  required. 

The  feed  systems  shown  in  figure  7 will, 
under  normal  conditions,  show  the  lowest  loss- 
es of  any  type  of  feed  system  since  the  cur- 
rents flowing  in  the  matching  network  are  the 
lowest  of  all  the  systems  commonly  used.  The 
"Folded  Element’’  match  shown  in  figure  7A 
and  the  "Yoke’’  match  shown  in  figure  7B  are 
the  most  satisfactory  electrically  of  all  stand- 
ard feed  methods.  However,  both  methods  re- 
quire the  extension  of  an  additional  conductor 
out  to  the  end  of  the  driven  element  as  a por- 
tion of  the  matching  system.  The  folded-ele- 
ment match  is  best  on  the  50-Mc.  band  and 


HANDBOOK 


Stacked  Yagi  Arrays  4 99 


1^0.2  X 0.  2 X — ii[^0.2  X — 1^0.2  X-w^j 


473^S^  450^N^  4S0N. 

fMC  ^N^FmC  ^\fMC 


I CAIN  ABOUT  1ft  OB 
WITH  a SECTIONS 


CAIN  ABOUT  17  DB 


Figure  6 

STACKED  YAGI  ARRAYS 

If  is  possible  to  attain  a relatively  large  amount  of  gain  over  a limited  bandwidth  with  stacked 
yagi  arrays.  The  two-section  array  at  (A)  will  give  a gain  of  about  12  db,  while  adding  a third 
section  will  bring  the  gain  up  ta  abaut  IS  db.  Adding  two  additional  parasitic  directars  to  each 
sectian,  as  at  fC)  will  bring  the  gain  up  to  about  17  db. 


higher  where  the  additional  section  of  tubing 
may  be  supported  below  the  main  radiator  ele- 
ment without  undue  difficulty.  The  yoke-match 
is  more  satisfactory  mechanically  on  the  28- 


Mc.  and  14-Mc.  bands  since  it  is  only  neces- 
sary to  suspend  a wire  below  the  driven  ele- 
ment proper.  The  wire  may  be  spaced  below 
the  self-supporting  element  by  means  of  several 


500  Rotary  Beams 


THE  RADIO 


© 


3-WIRE  MATCH 


Rfeeo  = 9 
Raad. 


Figure  7 
DATA  FOR 
FOLDED-ELEMENT 
MATCHING  SYSTEMS 


In  all  normal  applications  of 
the  data  given  the  main  ele- 
ment as  shown  is  the  driven 
element  of  a multi-element 
parasitic  array.  Directors  and 
reflectors  hove  not  been 
shown  for  the  sake  of  clarity. 


small  strips  of  polystyrene  which  have  been 
drilled  for  both  the  main  element  and  the  small 
wire  and  threaded  on  the  main  element. 

The  Folded-Element  The  calculation  of  the 

Match  Calculations  operating  conditions  of 

the  fol  ded- element 
matching  system  and  the  yoke  match,  as  shown 
in  figures  7A  and  7B  is  relatively  simple.  A 
selected  group  of  operating,  conditions  has 
been  shown  on  the  drawing  of  figure  7.  In  ap- 
plying the  system  it  is  only  necessary  to  mul- 
tiply the  ratio  of  feed  to  radiation  resistance 
(given  in  the  figures  to  the  right  of  the  sug- 
gested operating  dimensions  in  figure  7)  by 
the  radiation  resistance  of  the  antenna  system 
to  obtain  the  impedance  of  the  cable  to  be 
used  in  feeding  the  array.  Approximate  values 
of  radiation  resistance  for  a number  of  com- 
monly used  parasitic-element  arrays  are  given 
in  figure  5. 

As  an  example,  suppose  a 3-element  array 
with  0.15D-0.15R  spacing  between  elements  is 


to  be  fed  by  means  of  a 465-ohm  line  con- 
structed of  no.  12  wire  spaced  2 inches.  The 
approximate  radiation  resistance  of  such  an 
antenna  array  will  be  20  ohms.  Hence  we  need 
a ratio  of  impedance  step  up  of  23  to  obtain 
a match  between  the  characteristic  impedance 
of  the  transmission  line  and  the  radiation  re- 
sistance of  the  driven  element  of  the  antenna 
array.  Inspection  of  the  ratios  given  in  figure 
7 shows  that  the  fourth  set  of  dimensions 
given  under  figure  7B  will  give  a 24-to-l  step 
up,  which  is  sufficiently  close.  So  it  is  merely 
necessary  to  use  a 1-inch  diameter  driven  ele- 
ment with  a no.  8 wire  spaced  on  1 inch  centers 
CA  inch  below  the  outside  wall  of  the  1-inch 
tubing)  below  the  1-inch  element.  The  no.  8 
wire  is  broken  and  a 2-inch  insulator  placed 
in  the  center.  The  feed  line  then  carries  from 
this  insulator  down  to  the  transmitter.  The 
center  insulator  should  be  supported  rigidly 
from  the  1-inch  tube  so  that  the  spacing  be- 
tween the  piece  of  tubing  and  the  no.  8 wire 
will  be  accurately  maintained. 


HANDBOOK 


Matching  Systems  501 


Figure  8 

AVERAGE  DIMENSIOKS 
FOR  THE  DELTA  AND 
"T"  MATCH 


In  many  cases  it  will  be  desired  to  use  the 
folded-element  or  yoke  matching  system  with 
different  sizes  of  conductors  or  different  spac- 
ings  than  those  shown  in  figure  7.  Note,  then, 
that  the  impedance  transformation  ratio  of 
these  types  of  matching  systems  is  dependent 
both  upon  the  ratio  of  conductor  diameters  and 
upon  their  spacing.  The  following  equation 
has  been  given  by  Roberts  (liCA  Review,  June, 
1947)  for  the  determination  of  the  impedance 
transformation  when  using  different  diameters 
in  the  two  sections  of  a folded  element: 

/ z. 

Transformation  ratio  = ( ^ ^ 

V Z, 

In  this  equation  Zj  is  the  characteristic  im- 
pedance of  a line  made  up  of  the  smaller  of 
the  two  conductor  diameters  spaced  the  center- 
to-centet  distance  of  the  two  conductors  in  the 
antenna,  and  Zj  is  the  characteristic  imped- 
ance of  a line  made  up  of  two  conductors  the 
size  of  the  larger  of  the  two.  This  assumes 
that  the  feed  line  will  be  connected  in  series 
with  the  smaller  of  the  two  conductors  so  that 
an  impedance  step  up  of  greater  than  four  will 
be  obtained.  If  an  impedance  step  up  of  less 
than  four  is  desired,  the  feed  line  is  connected 
in  series  with  the  larger  of  the  two  conductors 
and  Zi  in  the  above  equation  becomes  the  im- 
pedance of  a hypothetical  line  made  up  of  the 
larger  of  the  two  conductors  and  Zj  is  made 
up  of  the  smaller.  The  folded  v-h-f  unipole  is 
an  example  where  the  transmission  line  is  con- 
nected in  series  with  the  larger  of  the  two 
conductors. 


The  conventional  3-wire  match  to  give  an 
impedance  ‘multiplication  of  9 and  the  5-wire 
match  to  give  a ratio  of  approximately  25  are 
shown  in  figures  7C  and  7D.  The  4-wire  match, 
not  shown,  will  give  an  impedance  transforma- 
tion ratio  of  approximately  16. 


The  Delta  Mutch  The  Delta  match  and  the 
and  T-Match  T-match  are  shown  in  figure 

8.  The  delta  match  has  been 
largely  superseded  by  the  newer  T-match,  how- 
ever both  these  systems  can  be  adjusted  to 
give  a low  value  of  SWR  on  50  to  600-ohm  bal- 
anced transmission  lines.  In  the  case  of  the 
systems  shown  it  will  be  necessary  to  make 
adjustments  in  the  tapping  distance  along  the 
driven  radiator  until  minimum  standing  waves 
on  the  antenna  transmission  line  are  obtained. 
Since  it  is  sometimes  impracticable  to  elim- 
inate completely  the  standing  waves  from  the 
antenna  transmission  line  when  using  these 
matching  systems,  it  is  common  practice  to 
cut  the  feed  line,  after  standing  waves  have 
been  reduced  to  a minimum,  to  a length  which 
will  give  satisfactory  loading  of  the  transmitter 
over  the  desired  frequency  range  of  operation. 

The  inherent  reactance  of  the  T-match  is 
tuned  out  by  the  use  of  two  identical  resonat- 
ing capacitors  in  series  with  each  leg  of  the 
T-tod.  These  capacitors  should  each  have  a 
maximum  capacity  of  8 pfiid.  per  meter  of  wave- 
length. Thus  for  20  meters,  each  capacitor 
should  have  a maximum  capacity  of  at  least 
160  fi/iid.  For  power  up  to  a kilowatt,  1000 
volt  spacing  of  the  capacitors  is  adequate. 


502  Rotary  Beams 


THE  RADIO 


DIRECT  FEED  WITH 
^ COAXIAL  CABLE 


(D  QUARTER-WAVE 
TRANSFORMER  FEED 


Figure  9 

ALTERNATE  FEED 
METHODS  WHERE  THE 
DRIVEN  ELEMENT  MAY 
BE  BROKEN  IN  THE 
CENTER 


These  capacitors  should  be  tuned  for  minimum 
SWR  on  the  transmission  line.  The  adjustment 
of  these  capacitors  should  be  made  at  the  same 
time  the  correct  setting  of  the  T-match  tods  is 
made  as  the  two  adjustments  tend  to  be  inter- 
locking. The  use  of  the  standing  wave  meter 
(described  in  Test  Equipment  chapter)  is 
recommended  for  making  these  adjustments  to 
the  T-match. 

Feed  Systems  Using  Four  methods  of  exciting 
a Driven  Element  the  driven  element  of  a 

with  Center  Feed  parasitic  array  are  shown 
in  figure  9.  The  system 
shown  at  (A)  has  proven  to  be  quite  satisfac- 
tory in  the  case  of  an  antenna-reflector  two- 
element  array  or  in  the  case  of  a three-element 
array  with  0.2  to  0.25  wavelength  spacing  be- 
tween the  elements  of  the  antenna  system.  The 
feed-point  impedance  of  the  center  of  the  driven 
element  is  close  enough  to  the  characteristic 
impedance  of  the  52-ohm  coaxial  cable  so  that 


the  standing-wave  ratio  on  the  52-ohm  coaxial 
cable  is  less  than  2-to-l.  (B)  shows  an  arrange- 
ment for  feeding  an  array  with  a broken  driven 
element  from  an  open-wire  line  with  the  aid  of 
a quartet-wave  matching  transformer.  With  465- 
ohm  line  from  the  transmitter  to  the  antenna 
this  system  will  give  a close  match  to  a 12- 
ohm  impedance  at  the  center  of  the  driven  ele- 
ment. (C)  shows  an  arrangement  which  uses  an 
untuned  transformer  with  lumped  inductance 
for  matching  the  transmission  line  to  the  cen- 
ter impedance  of  the  driven  element. 

Rotary  Link  In  many  cases  it  is  desirable  to 

Coupling  be  able  to  allow  the  antenna  ar- 

ray to  rotate  continuously  without 
regard  to  snarling  of  the  feed  line.  If  this  is  to 
be  done  some  sort  of  slip  rings  or  rotary  joint 
must  be  made  in  the  feed  line.  One  relatively 
simple  methodof  allowing  unrestrained  rotation 
of  .the  antenna  is  to  use  the  method  of  rotary 
link  coupling  shown  in  figure  9D.  The  two  cou- 


HANDBOOK 


The  Gamma  Match  503 


Figure  10 

THE  GAMMA  MATCHING  SYSTEM 
See  text  for  details  of  resonating  copacifor 


•FLAT-  LINE  RESONANT 

SWR*1.0  SECTION 

A — A 


t \ / \ 

TO  TRANSMITTER 

1 ANYANTENHAZ 

1 SIMPLE  OR  COMPLEX 

s. 

MATCHINO  STUB 

Figure  1 1 

IMPEDANCE  MATCHING  WITH  A CLOSED 
STUB  ON  A TWO  WIRE  TRANSMISSION 
LINE 


pling  rings  are  10  inches  in  diameter  and  are 
usually  constructed  of  ^-inch  copper  tubing 
supported  one  from  the  rotating  structure  and 
one  from  the  fixed  sttucture  by  means  of  stand- 
off insulators.  The  capacitor  C in  figure  9D  is 
adjusted,  after  the  antenna  has  been  tuned,  for 
minimum  standing-wave  ratio  on  the  antenna 
transmission  line.  The  dimensions  shown  will 
allow  operation  with  either  14-Mc.  or  28-Mc. 
elements,  with  appropriate  adjustment  of  the 
capacitor  C.  The  rings  must  of  course  be  paral- 
lel and  must  lie  in  a plane  normal  to  the  axis 
of  rotation  of  the  rotating  structure. 

The  Gamma  Match  The  use  of  coaxial  cable 
to  feed  the  driven  element 
of  a yagi  array  is  becoming  increasingly  popu- 
lar. One  reason  for  this  increased  popularity 
lies  in  the  fact  that  the  TVI-reduction  problem 
is  simplified  when  coaxial  feed  line  is  used 
from  the  transmitter  to  the  antenna  system. 
Radiation  from  the  feed  line  is  minimized  when 
coaxial  cable  is  used,  since  the  outer  conduc- 
tor of  the  line  may  be  grounded  at  several 
points  throughout  its  length  and  since  the  in- 
tense field  is  entirely  confined  within  the  out- 
er conductor  of  the  coaxial  cable.  Other  ad- 
vantages of  coaxial  cable  as  the  antenna  feed 
line  lie  in  the  fact  that  coaxial  cable  may  be 
run  within  the  sttucture  of  a building  without 
danger,  or  the  cable  may  be  run  underground 
without  disturbing  its  operation.  Also,  trans- 
mitting-type  low-pass  filters  for  52  ohm  imped- 
ance are  more  widely  available  and  ate  less 
expensive  than  equivalent  filters  for  two-wire 
line. 

The  gamma-match  is  illustrated  in  figure  10, 
and  may  be  looked  upon  as  one-half  of  a T- 
match.  One  resonating  capacitor  is  used, 
placed  in  series  with  the  gamma  tod.  The  ca- 
pacitor should  have  a capacity  of  7 g/tfd.  per 
meter  of  wavelength.  For  15-meter  operation 
the  capacitor  should  have  a maximum  capacity 
of  105  fififd.  The  length  of  the  gamma  rod  deter- 
mines the  impedance  transformation  between 


the  transmission  line  and  the  driven  element 
of  the  array,  and  the  gamma  capacitor  tunes 
out  the  inductance  of  the  gamma  rod.  By  ad- 
justment of  the  length  of  the  gamma  rod,  and 
the  setting  of  the  gamma  capacitor,  the  SWR 
on  the  coaxial  line  may  be  brought  to  a very 
low  value  at  the  chosen  operating  frequency. 
The  use  of  an  Antennascope,  described  in  the 
Test  Equipment  chapter  is  recommended  for 
precise  adjustment  of  the  gamma  match. 

The  Matching  Stub  If  an  open-wire  line  is 
used  to  fe.ed  a low  imped- 
ance radiator,  a section  of  the  transmission 
line  may  be  employed  as  a matching  stub  as 
shown  in  figure  11.  The  matching  stub  can 
transform  any  complex  impedance  to  the  char- 
acteristic impedance  of  the  transmission  line. 
While  it  is  possible  to  obtain  a perfect  match 
and  good  performance  with  either  an  open  stub 
or  a shotted  one  by  observing  appropriate  di- 
mensions, a shotted  stub  is  much  more  readily 
adjusted.  Therefore,  the  following  discussion 
will  be  confined  to  the  problem  of  using  a 
closed  stub  to  match  a low  impedance  load  to 
a high  impedance  transmission  line. 

If  the  transmission  line  is  so  elevated  that 
adjustment  of  a "fundamental”  shorted  stub 
cannot  be  accomplished  easily  from  the  ground, 
then  the  stub  length  may  be  increased  by  ex- 
actly one  or  two  electrical  half  wavelengths, 
without  appreciably  affecting  its  operation. 

While  the  correct  position  of  the  shorting 
bar  and  the  point  of  attachment  of  the  stub  to 
the  line  ctui  be  determined  entirely  by  experi- 
mental methods,  the  fact  that  the  two  adjust- 
ments are  interdependent  or  interlocking  makes 
such  a cut-and-try  procedure  a tedious  one. 
Much  time  can  be  saved  by  determining  the  ap- 
proximate adjustments  requited  by  reference  to 
a chart  such  as  figure  12  and  using  them  as  a 
starter.  Usually  only  a slight  "touching  up” 
will  produce  a perfect  match  and  flat  line. 

In  order  to  utilize  figure  12,  it  is  first  neces- 
sary to  locate  accurately  a voltage  node  or 
current  node  on  the  line  in  the  vicinity  that 


504  Rotary  Beams 


THE  RADIO 


SWR 


Figure  12 

SHORTED  STUB  LENGTH  AND  POSITION 
CHART 


From  the  standing  wave  ratio  and  current  or 
voltage  null  position  it  is  possible  to  deter' 
mine  the  theoretically  correct  length  and 
position  o4*  a shorted  stub'  In  sctual  prac 
tice  a slight  discrepancy  usually  will  be 
found  between  the  theoretical  and  fhe  ex- 
perlmentally  optimized  dimensions;  therefore 
It  may  be  necessary  to  touch  up  fhe  di' 
mensions  after  using  the  above  data  as  a 
starting  point. 


has  been  decided  upon  for  the  stub,  and  also 
to  determine  the  SWR. 

Stub  adjustment  becomes  more  critical  as 
the  SWR  increases,  and  under  conditions  of 
high  SWR  the  current  and  voltage  nulls  are 
more  sharply  defined  than  the  current  and  volt- 
age maxima,  or  loops.  Therefore,  it  is  best  to 
locate  either  a current  null  or  voltage  null,  de- 
pending upon  whether  a current  indicating  de- 
vice or  a voltage  indicating  device  is  used  to 
check  the  standing  wave  pattern. 

The  SWR  is  determined  by  means  of  a "di- 
rectional coupler,’’  or  by  noting  the  ratio  of 
^max  Rmin  ^max  Imin  read  on  an 
indicating  device. 

It  is  assumed  that  the  characteristic  imped- 
ance of  the  section  of  line  used  as  a stub  is 
the  same  as  that  of  the  transmission  line  prop- 
er. It  is  preferable  to  have  the  stub  section 
identical  to  the  line  physically  as  well  as 
electrically. 


25-5  Unidirectional 

Driven  Arrays 


Three  types  of  unidirectional  driven  arrays 
are  illustrated  in  figure  13.  The  array  shown 
in  figure  13A  is  an  end-fire  system  which  may 


DIRECTIONAL 


Figure  13 

UNIDIRECTIONAL  ALL-DRIVEN  ARRAYS 

A unJdireetionaf  a/l-driven  end^fire  array  is 
shown  at  (A).  (B)  shows  on  array  with  two 
ha!f  waves  in  phase  with  driven  reflectors, 
A Lazy^H  array  with  driven  reflectors  is 
shown  at  (C),  Note  that  the  directivity  is 
through  the  elements  with  the  greatest  total 
feed’line  length  in  arrays  such  as  shown  at 
(B)  and  (Cl 


be  used  in  place  of  a parasitic  array  of  similar 
dimensions  when  greater  frequency  coverage 
than  is  available  with  theyagi  type  is  desired. 
Figure  13B  is  a combination  end-fire  and  co- 


HANDBOOK 


Driven  Arrays  5 05 


linear  system  which  will  give  approximately 
the  same  gain  as  the  system  of  figure  13A,  but 
which  requires  less  boom  length  and  greater 
total  element  length.  Figure  13C  illustrates 
the  familiar  lazy-H  with  driven  reflectors  (or 
directors,  depending  upon  the  point  of  view) 
in  a combination  which  will  show  wide  band- 
width with  a considerable  amount  of  forward 
gain  and  good  front-to-back  ratio  over  the  en- 
tire frequency  coverage. 

Unidirectional  Stacked  Three  practicable 

Broadside  Arrays  types  of  unidirectional 

stacked  broadside  ar- 
rays are  shown  in  figure  14.  The  first  type, 
shown  at  figure  14A,  is  the  simple  *'lazy  H’* 
type  of  antenna  with  parasitic  reflectors  for 
each  element.  (B)  shows  a simpler  antenna  ar- 
ray with  a pair  of  folded  dipoles  spaced  one- 
half  wave  vertically,  operating  with  reflectors. 
In  figure  14C  is  shown  a more  complex  array 
with  six  half  waves  and  six  reflectors  which 
will  give  a very  worthwhile  amount  of  gain. 

In  all  three  of  the  antenna  arrays  shown  the 
spacing  between  the  driven  elements  and  the 
reflectors  has  been  shown  as  one-quarter  wave- 
length. This  has  been  done  to  eliminate  the 
requirement  for  tuning  of  the  reflector,  as  a 
result  of  the  fact  that  a half-wave  element 
spaced  exactly  one-quarter  wave  from  a driven 
element  will  make  a unidirectional  array  when 
both  elements  are  the  same  length.  Using  this 
procedure  will  give  a gain  of  3 db  with  the  re- 
flectors over  the  gain  without  the  reflectors, 
with  only  a moderate  decrease  in  the  radiation 
resistance  of  the  driven  element.  Actually, 
the  radiation  resistance  of  a half-wave  dipole 
goes  down  from  73  ohms  to  60  ohms  when  an 
identical  half-wave  element  is  placed  one- 
quarter  wave  behind  it. 

A very  slight  increase  in  gain  for  the  entire 
array  (about  1 db)  mt^  be  obtained  at  the  ex- 
pense of  lowered  radiation  resistance,  the  ne- 
cessity for  tuning  the  reflectors,  and  decreased 
bandwidth  by  placing  the  reflectors  0.15  wave- 
length behind  the  driven  elements  and  making 
them  somewhat  longer  than  the  driven  elements. 
The  radiation  resistance  of  each  element  will 
drop  approximately  to  one-half  the  value  ob- 
tained with  untuned  half-wave  reflectors  spaced 
one-quarter  wave  behind  the  driven  elements. 

Antenna  arrays  of  the  type  shown  in  figure 
14  require  the  use  of  some  sort  of  lattice  work 
for  the  supporting  structure  since  the  arrays 
occupy  appreciable  distance  in  space  in  all 
three  planes. 

Feed  Methods  The  requirements  for  the  feed 
systems  for  antenna  arrays  of 
the  type  shown  in  figure  14  are  less  critical 
than  those  for  the  close-spaced  parasitic  ar- 
rays shown  in  the  previous  section.  This  is  a 


natural  result  of  the  fact  that  a larger  number 
of  the  radiating  elements  are  directly  fed  with 
energy,  and  of  the  fact  that  the  effective  radia- 
tion resistance  of  each  of  the  driven  elements 
of  the  array  is  much  higher  than  the  feed-point 
resistance  of  a parasitic  array.  As  a conse- 
quence of  this  fact,  arrays  of  the  type  shown 
in  figure  14  can  be  expected  to  cover  a some- 
what greater  frequency  band  for  a specified 
value  of  standing-wave  ratio  than  the  parasitic 
type  of  array. 

In  most  cases  a simple  open-wire  line  may 
be  coupled  to  the  feed  point  of  the  array  with- 
out any  matching  system.  The  standing-wave 
ratio  with  such  a system  of  feed  will  often  be 
less  than  2-to-l.  However,  if  a more  accurate 
match  between  the  antenna  transmission  line 
and  the  array  is  desired  a conventional  quar- 
ter-wave stub,  or  a quarter-wave  matching 
transformer  of  £q>propriate  impedance,  may  be 
used  to  obtain  a low  standing-wave  ratio. 


25-6  Bi-Direclional 

Rotaloble  Arrays 

The  bi-directional  type  of  array  is  sometimes 
used  on  the  28-Mc.  and  50-Mc.  bands  where 
signals  ate  likely  to  be  coming  from  only  one 
general  direction  at  a time.  Hence  the  sacri- 
fice of  discrimination  against  signals  arriving 
from  the  opposite  direction  is  likely  to  be  of 
little  disadvantage.  Figure  15  shows  two  gen- 
eral types  of  bi-directional  arrays.  The  flat- 
top  beam,  which  has  been  described  in  detail 
earlier,  is  well  adapted  to  installation  atop  a 
rotating  structure.  When  self-supporting  ele- 
ments are  used  in  the  flat-top  beam  the  prob- 
lem of  losses  due  to  insulators  at  the  ends  of 
the  elements  is  somewhat  reduced.  With  a 
single-section  flat-top  beam  a gain  of  approx- 
imately 4 db  can  be  expected,  and  with  two 
sections  a gain  of  approximately  6 db  can  be 
obtained. 

Another  type  of  bi-directional  array  which 
has  seen  less  use  than  it  deserves  is  shown 
in  figure  15B.  This  type  of  antenna  system  has 
a relatively  broad  azimuth  or  horizontal  beam, 
being  capable  of  receiving  signals  with  little 
diminution  in  strength  over  approximately  40°, 
but  it  has  a quite  sharp  elevation  pattern  since 
substantially  all  radiation  is  concentrated  at 
the  lower  angles  of  radiation  if  mote  than  a 
total  of  four  elements  is  used  in  the  antenna 
system.  Figure  15B  gives  the  approximate  gain 
over  a half-wave  dipole  at  the  height  of  the 
center  of  the  array  which  can  be  expected.  Al- 
so shown  in  this  figure  is  a type  of  "rotating 
mast”  structure  which  is  well  suited  to  rota- 
tion of  this  type  of  array. 


5 06  Rotary  Beams 


THE  RADIO 


"LAZY  H*  WITH  REFLECTOR 


GAIN  APPROX.  9 OB 


® 

BROADSIDE  HALF-WAVES 
WITH  REFLECTORS 


GAIN  APPROX.  7 DB 


© 

“TWO  OVER  TWO  OVER  TWO" 
WITH  REFLECTORS 


SAIN  APPROX.  II. » 06 


Figure  14 

BROADSIDE  ARRAYS 
WITH  PARASITIC 
REFLECTORS 

The  apparent  gain  of  the  or- 
rays  Illustrated  will  be  greats 
er  than  the  values  given  due 
to  concentration  of  the  radi* 
ated  signal  at  the  lower  ele- 
vation angles. 


If  six  or  more  elements  are  used  in  the  type 
of  array  shown  in  figure  15B  no  matching  sec- 
tion will  be  required  between  the  antenna  trans- 
mission line  and  the  feed  point  of  the  antenna. 
When  only  four  elements  are  used  the  antenna 
is  the  familiar  "lazy  H”  and  a quarter- wave 
stub  should  be  used  for  feeding  from  the  an- 
tenna transmission  line  to  the  feed  point  of 
the  antenna  system. 

If  desired,  and  if  mechanical  considerations 
permit,  the  gain  of  the  arrays  shown  in  figure 
15B  may  be  increased  by  3 db  by  placing  a 
half-wave  reflector  behind  each  of  the  ele- 
ments at  a spacing  of  one-quarter  wave.  The 
array  then  becomes  essentially  the  same  as 
that  shown  in  figure  14C  and  the  same  con- 
siderations in  regard  to  reflector  spacing  and 


tuning  will  apply.  However,  the  factor  that  a 
bi-directional  array  need  be  rotated  through  an 
angle  of  less  than  180°  should  be  considered 
in  this  connection. 


25-7  Construction  of 

Rotatable  Arrays 

A considerable  amount  of  ingenuity  may  be 
exercised  in  the  construction  of  the  supporting 
structure  for  a rotatable  array.  Every  person 
has  his  own  ideas  as  to  the  best  method  of 
construction.  Often  the  most  practicable  meth- 
od of  construction  will  be  dictated  by  the 


HANDBOOK 


Bi-directional  Arrays  507 


Figure  15 

TWO  GENERAL  TYPES 
OF  BI-DIRECTIONAL 
ARRAYS 

Average  gain  figures  are  giv- 
en for  both  the  flat-top  beam 
type  of  array  and  for  the 
broadside-cotinear  array  with 
different  numbers  af  elements* 


availability  of  certain  types  of  constructional 
materials.  But  in  any  event  be  sure  that  sound 
mechanical  engineering  principles  are  used  in 
the  design  of  the  supporting  structure.  There 
are  few  things  quite  as  discouraging  as  the 
picking  up  of  pieces,  repairing  of  the  roof,  etc., 
when  a newly  constructed  rotary  comes  down 
in  the  first  strong  wind.  If  the  principles  of 
mechanical  engineering  are  understood  it  is 
wise  to  calculate  the  loads  and  torques  which 
will  exist  in  the  various  members  of  the  struc- 
ture with  the  highest  wind  velocity  which  may 
be  expected  in  the  locality  of  the  installation. 
If  this  is  not  possible  it  will  usually  be  worth 
the  time  and  effort  to  look  up  a friend  who 
understands  these  principles. 

Radioting  One  thing  more  or  less  standard 
Elements  about  the  construction  of  rotatable 

antenna  arrays  is  the  use  of  dural 
tubing  for  the  self-supporting  elements.  Other 
materials  may  be  used  but  an  alloy  known  as 


24ST  has  proven  over  a period  of  time  to  be 
quite  satisfactory.  Copper  tubing  is  too  heavy 
for  a given  strength,  and  steel  tubing,  unless 
copper  plated,  is  likely  to  add  an  undesirably 
large  loss  resistance  to  the  array.  Also,  steel 
tubing,  even  when  plated,  is  not  likely  to 
withstand  salt  atmosphere  such  as  encountered 
along  the  seashore  for  a satisfactory  period 
of  time.  Do  not  use  a soft  aluminum  alloy  for 
the  elements  unless  they  will  be  quite  short; 
24ST  is  a hard  alloy  and  is  best  although  there 
are  several  other  alloys  ending  in  "ST”  which 
will  be  found  to  be  satisfactory.  Do  not  use 
an  alloy  ending  in  "SO”  or  "S”  in  a position 
in  the  array  where  structural  strength  is  im- 
portant, since  these  letters  designate  a metal 
which  has  not  been  heat  treated  for  strength 
and  rigidity.  However,  these  softer  alloys,  and 
aluminum  electrical  conduit,  may  be  used  for 
short  radiating  elements  such  as  would  be 
used  for  the  50-Mc.  band  or  as  interconnecting 
conductors  in  a stacked  array. 


5 08  Rotary  Beams 


THE  RADIO 


LINE  OF 
ELEMENT 


@ 


r — SHIM  JOINT  WITH  THIN 
\ STRIPS  OF  ALUMINUM 
\ IF  NECESSARY. 


12' CENTER  SECTION 


TYPICAL  ELEMENT 


Figure  16 

3-ELEMENT  “PLUMBER'S 
DELIGHT"  ANTENNA  ARRAY 

All-metal  configuration  permits  rugged, 
light  assembly.  Joints  are  made  with 
U-bolts  and  metal  plates  for  maximum 
rigidity. 


“Plumber's  Delight"  It  is  characteristic  of 

Construction  conventional  type  of 

multi-element  parasitic 
array  such  as  discussed  previously  and  out- 
lined that  the  centers  of  all  the  elements  are  at 
zero  r-f  potential  with  respect  to  ground.  It  is 
therefore  possible  to  use  a metallic  structure 
without  insulators  for  supporting  the  various 
elements  of  the  array.  A typical  three  element 
array  of  this  type  is  shown  in  figure  16.  In  this 
particular  array,  U-bolts  and  metal  plates  have 
been  employed  to  fasten  the  elements  to  the 
boom.  The  elements  are  made  of  telescoping 
sections  of  aluminum  tubing.  The  tips  of  the 
inner  sectionsof  tubing  are  split,  and  a tubing 
clamp  is  slipped  over  the  joint,  as  shown  in 
the  drawing.  Before  assembly  of  the  joint,  the 
mating  pieces  of  aluminum  are  given  a thin 
coat  of  Penetrox-A  compound.  (This  anti- 
oxidizing paste  is  manufactured  by  Burndy  Co,, 
Norwalk,  Conn,  and  is  distributed  by  the 
General  Electric  Supply  Co.)  When  the  tubes 
are  telesaoped  and  the  clamp  is  tightened,  an 
air-tight  seal  is  produced,  reducing  corrosion 
to  a minimum. 

The  boom  of  the  parasitic  array  may  be  made 
from  two  or  three  sections  of  steel  TV  mast, 
or  it  may  be  made  of  a single  section  of  alumi- 
num irrigation  pipe.  This  pipe  is  made  by 
Reynolds  Aluminum  Co.,  and  others,  and  may 
often  be  purchased  via  the  Sears,  Roebuck  Co. 
mail-order  department.  Three  inch  pipe  may  be 


OXEN-YOKE  CLAMP 


Figure  17 

(A)  OXEN-YOKE  CLAMP  IS  DESIGNED  FOR 
ALL  METAL  ASSEMBLY 

(B>  ALTERNATIVE  WOODEN  SUPPORTING 
ARRANGEMENT 

A wooden  ladder  may  be  used  to  support  a 
10  or  15  meter  array. 

used  for  the  10  and  15  meter  antennas,  and  the 
huskier  four  inch  pipe  should  be  used  for  a 
20  meter  beam. 

Automobile  muffler  clamps  can  often  be 
used  to  affix  the  elements  to  the  support 
plates.  Larger  clamps  of  this  type  will  fasten 
the  plates  to  the  boom.  In  most  cases,  the 
muffler  clamps  are  untreated,  and  they  should 
be  given  one  or  two  coats  of  rust-proof  paint 
to  protect  them  from  inclement  weather.  All 
bolts,  nuts,  and  washers  used  in  the  assembly 
of  the  array  should  be  of  the  plated  variety  to 
reduce  corrosion  and  rust. 

An  alternative  assembly  is  to  employ  the 
"Oxen  Yoke"  type  of  clamps,  shown  in  figure 
17.  These  light-weight  aluminum  fittings  are 
obtainable  from  the  Continental  Electronics 
and  Sound  Co,,  Dayton,  27,  Ohio,  and  are 
available  in  a wide  range  of  sizes. 

If  it  is  desired  to  use  a split  driven  element 
for  a balanced  feed  system,  it  is  necessary  to 
insulate  the  element  from  the  supporting  struc- 
ture of  the  antenna.  The  element  should  be 
severed  at  the  center,  and  the  two  halves 


HANDBOOK 


T uning  the  Array  509 


driven  onto  a wooden  dowel,  as  shown  in 
figure  17B.  The  element  may  then  be  mounted 
upon  an  aluminum  support  plate  by  means  of 
four  ceramic  insulators.  Metal  based  insulators, 
such  as  the  Johnson  135-67  are  recommended, 
since  the  all-ceramic  types  may  break  at  the 
mounting  holes  when  the  array  is  subjected 
to  heavy  winds. 

25-8  Tuning  the  Array 

Although  satisfactory  results  may  be  ob- 
tained by  pre-cutting  the  antenna  array  to  the 
dimensions  given  earlier  in  this  chapter,  the 
occasion  might  arise  when  it  is  desired  to 
make  a check  on  the  operation  of  the  antenna 
before  calling  the  job  complete. 

The  process  of  tuning  an  array  may  fairly 
satisfactorily  be  divided  into  two  mote  or  less 
distinct  steps:  the  actual  tuning  of  the  array 
for  best  front-to-back  ratio  or  for  maximum  for- 
ward gain,  and  the  project  of  obtaining  the 
best  possible  impedance  match  between  the 
antenna  transmission  line  and  the  feed  point 
of  the  array. 

Tuning  the  The  actual  tuning  of  the  array 
Array  Proper  for  best  front-to-back  ratio  or 
maximum  forward  gain  may  best 
be  accomplished  with  the  aid  of  a low-power 
transmitter  feeding  a dipole  antenna  (polarized 
the  stune  as  the  array  being  tuned)  at  least 
four  or  five  wavelengths  away  from  the  anten- 
na being  tuned  and  located  at  the  same  eleva- 
tion as  that  of  the  antenna  under  test.  A cali- 
brated field-strength  meter  of  the  remote-indi- 
cating type  is  then  coupled  to  the  feed  point 
of  the  antenna  array  being  tuned.  The  trans- 
missions from  the  portable  transmitter  should 
be  made  as  short  as  possible  and  the  call  sign 
of  the  station  making  the  test  should  be  trans- 
mitted at  least  every  ten  minutes. 

It  is,  of  course,  possible  to  tune  an  array 
with  the  receiver  connected  to  it  and  with  a 
station  a mile  or  two  away  making  transmis- 
sions on  your  request.  But  this  method  is  more 
cumbersome  and  is  not  likely  to  give  complete 
satisfaction.  It  is  also  possible  to  carry  out 
the  tuning  process  with  the  transmitter  con- 
nected to  the  array  and  with  the  field- strength 
meter  connected  to  the  remote  dipole  antenna. 
In  this  event  the  indicating  instrument  of  the 
remote-indicating  field-strength  meter  should 
be  visible  from  the  position  where  the  elements 
are  being  tuned.  However,  when  the  array  is 
being  tuned  with  the  transmitter  connected  to 
it  there  is  always  the  problem  of  making  con- 
tinual adjustments  to  the  transmitter  so  that  a 
constant  amount  of  power  will  be  fed  to  the 
array  under  test.  Also,  if  you  use  this  system. 


use  very  low  power  (5  or  10  watts  of  power  is 
usually  sufficient)  and  make  sure  that  the  an- 
tenna transmission  line  is  effectively  grounded 
as  far  as  d-c  plate  voltage  is  concerned.  The 
use  of  the  method  described  in  the  previous 
paragraph  of  course  eliminates  these  problems. 

One  satisfactory  method  for  tuning  the  array 
proper,  assuming  that  it  is  a system  with  sev- 
eral parasitic  elements,  is  to  set  the  directors 
to  the  dimensions  given  in  figure  5 and  then 
to  adjust  the  reflector  for  maximum  forward 
signal.  Then  the  first  director  should  be  varied 
in  length  until  maximum  forward  signal  is  ob- 
tained, and  so  on  if  additional  directors  are 
used.  Then  the  array  may  be  reversed  in  direc- 
tion and  the  reflector  adjusted  for  best  front- 
to-back  ratio.  Subsequent  small  adjustments 
may  then  be  made  in  both  the  directors  and  the 
reflector  for  best  forward  signal  with  a reason- 
able ratio  of  front-to-back  signal.  The  adjust- 
ments in  the  directors  and  the  reflector  will 
be  found  to  be  interdependent  to  a certain  de- 
gree, but  if  small  adjustments  are  made  after 
the  preliminary  tuning  process  a satisfactory 
set  of  adjustments  for  maximum  performance 
will  be  obtained.  It  is  usually  best  to  make 
the  end  sections  of  the  elements  smaller  in 
diameter  so  that  they  will  slip  inside  the  larger 
tubing  sections.  The  smaller  sliding  sections 
may  be  clamped  inside  the  larger  main  sec- 
tions. 

In  making  the  adjustments  described,  it  is 
best  to  have  the  rectifying  element  of  the  re- 
mote-indicating field- strength  meter  directly 
at  the  feed  point  of  the  array,  with  a tesistor 
at  the  feed  point  of  the  estimated  value  of 
feed-point  impedance  for  the  array. 

Matching  to  the  The  problem  of  matching  the 
Antenna  Trans-  impedance  of  the  antenna 
mission  Line  transmission  line  to  the  array 
is  much  simplified  if  the  pto- 


Figure  18 

ADJUSTMENT  OF  GAMMA  MATCH  BY  USE 
OF  ANTENNASCOPE  AND  GRID-DIP 
METER 


IMPORTANT  NOTE 

BE  SURE  THAT  THE  PIPES  FIT 
WITHOUT  BINDING  AS  THERE 
IS  SOME  VARIATION  IN  DIA- 
METERS OF  THE  SIZES  USED. 


TWO  3/8'  OR  1/2'  MACHINE  . 
SCREWS  ABOUT  1/2*  LONG 
THROUGH  BEARING  PIECE 
AND  TAPPED  INT01'  PIPE. 


THE  STATIONARY  PIPE  CAN  WEATHER-TK 

BE  SECURELY  BRACED  TO  I ZED  IRON  R> 

ROOF  PARTS  BY  BLOCKS  I > 

AND  U- BOLTS —li--  -sL^—  


WEATHER-TIGHT  GALVAN- 
IZED IRON  RAIN  SHIELD 


-1'  IRON  PIPE 
LENGTH  TO  FIT 
AS  PER  OTHER 
DRAWINGS 


-ABOUT  2'  OF  1 1/4'  PIPE. 
THIS  15  THE  BEARING  THAT 
CARRIES  THE  LOAD. 


«LONG  THREAD*  SO  THAT  1 1/4'  PIPE 
■^EXTENDS  ABOUT  1/2' ABOVE  FLANGE. 


2 OF  THE  GUY  WIRES  SHOWN 


-1  1/4'  IRON  PIPE 


LENGTH  TO  FIT 
AS  PER  OTHER 
DRAWINGS. 


PLANK  OVER  SEVERAL 
JOISTS  TO  DISTRIBUTE 
THE  WEIGHT 


iv  RG8U  OR  SMALLER 
V ANTENNA  FEEDER 
' MAY  BE  RUN  THRU  PI  PE 


1'PIPE  LOWERS  TO 
INSTALL  ANTENNA  OR 
ADJUST 


REMOVABLE  EXTEN- 
SION MAY  BE  USED 
TO  LOWER  THE  WHEEL- 


PITCHED  ROOF  INSTALLATION 


FLAT  OR  SHED  ROOF  INSTALLATION 


510  Rotary  Beams  THE  RADIO 


HANDBOOK 


Tuning  the  Array  511 


cess  of  tuning  the  array  is  made  a substan- 
tially separate  process  as  just  described.  After 
the  tuning  operation  is  complete,  the  resonant 
frequency  of  the  driven  element  of  the  antenna 
should  be  checked,  directly  at  the  center  of 
the  driven  element  if  practicable,  with  a grid- 
dip  meter.  It  is  important  that  the  resonant  fre- 
quency of  the  antenna  be  at  the  center  of  the 
frequency  band  to  be  coveted.  If  the  resonant 
frequency  is  found  to  be  much  different  from 
the  desired  frequency,  the  length  of  the  driven 
element  of  the  array  should  be  altered  until 
this  condition  exists.  A relatively  small  change 
in  the  length  of  the  driven  element  will  have 
only  a second  order  effect  on  the  tuning  of  the 
parasitic  elements  of  the  array.  Hence,  a mod- 
erate change  in  the  length  of  the  driven  ele- 
ment may  be  made  without  repeating  the  tuning 
process  for  the  parasitic  elements. 

When  the  tesonant  frequency  of  the  antenna 
system  is  correct,  the  antenna  transmission 
line,  with  impedance-matching  device  or  net- 
work between  the  line  and  antenna  feed  point, 
is  then  attached  to  the  array  and  coupled  to  a 
low-power  exciter  unit  or  transmitter.  Then, 
preferably,  a standing-wave  meter  is  connected 
in  series  with  the  antenna  transmission  line 
at  a point  relatively  much  more  close  to  the 
transmitter  than  to  the  antenna.  However,  for 
best  indication  there  should  be  10  to  15  feet 
of  line  between  the  transmitter  and  the  stand- 
ing-wave meter.  If  a standing-wave  meter  is 
not  available  the  standing-wave  ratio  may  be 
checked  approximately  by  means  of  a neon 
lamp  or  a short  fluorescent  tube  if  twin  trans- 
mission line  is  being  used,  or  it  may  be  check- 
ed with  a thermomilliammeter  and  a loop,  a 
neon  lamp,  or  an  r-f  ammeter  and  a pair  of 
clips  spaced  a fixed  distance  for  clipping  onto 
one  wire  of  a two-wire  open  line. 

If  the  standing-wave  ratio  is  below  1.5  to  1 


it  is  satisfactory  to  leave  the  installation  as 
it  is.  If  the  ratio  is  greater  than  this  range  it 
will  be  best  when  twin  line  or  coaxial  line  is 
being  used,  and  advisable  with  open-wire  line, 
to  attempt  to  decrease  the  s.w.t. 

It  must  be  remembered  that  no  adjustments 
made  at  the  transmitter  end  of  the  transmission 
line  will  alter  the  SW'R  on  the  line.  All  adjust- 
ments to  bettet  the  SW'R  must  be  made  at  the 
antenna  endof  the  line  and  to  the  device  which 
performs  the  impedance  ttansfotmation  neces- 
sary to  match  the  characteristic  impedance  of 
the  antenna  to  that  of  the  transmission  line. 

Before  any  adjustments  to  the  matching  sys- 
tem are  made,  the  resonant  frequency  of  the 
driven  element  must  be  ascertained,  as  ex- 
plained previously.  If  all  adjustments  to  cor- 
rect impedance  mismatch  are  made  at  this  fre- 
quency, the  problem  of  reactance  termination 
of  the  transmission  line  is  eliminated,  greatly 
simplifying  the  problem.  The  following  steps 
should  be  taken  to  adjust  the  impedance  trans- 
formation; 

1.  The  output  impedance  of  the  matching 
device  should  be  measured.  An  Antenna- 
scope  and  a grid-dip  oscillator  are  re- 
quired for  this  step.  The  Antennascope 
is  connected  to  the  output  terminals  of 
the  matching  device.  If  the  driven  element 
is  a folded  dipole,  the  Antennascope 
connects  directly  to  the  split  section  of 
the  dipole.  If  a gamma  match  or  T-match 
are  used,  the  Antennascope  connects  to 
the  transmission-line  end  of  the  device. 
If  a Q-section  is  used,  the  Antennascope 
connects  to  the  bottom  end  of  the  sec- 
tion. The  grid-dip  oscillator  is  coupled 
to  the  input  terminals  of  the  Antenna- 
scope as  shown  in  figure  18. 

2.  The  grid-dip  oscillator  is  tuned  to  the 
resonant  frequency  of  the  antenna,  which 


Figure  19 

ALL-PIPE  ROTATING  MAST  STRUCTURE  FOR  ROOF  INSTALLATION 

An  installation  suitable  far  a building  yritb  a pitched  roof  is  shown  at  (A).  At  (B)  is  shown  a 
similar  installation  for  a flat  or  shed  roof.  The  arrangement  as  shown  is  strong  enough  to  sup* 
port  a I ightweight  3-element  28-Mc,  array  and  o light  3-element  50-Mc,  array  above  the  28-Mc, 
array  on  the  end  of  a 4-foot  length  of  ’/l-inch  pipe. 

The  lengths  of  pipe  shown  were  chosen  so  that  when  the  system  is  in  the  lowered  position  one 
can  stand  on  a household  ladder  and  put  the  beam  in  position  atop  the  rotating  pipe.  The  lengths 
may  safely  be  revised  upward  somewhat  if  the  array  is  of  a particularly  lightweight  design  with 

low  wind  resistance. 

Just  before  the  mast  is  installed  It  is  a good  idea  to  give  the  rotating  pipe  a good  smearing  of 
cup  grease  or  waterproof  pump  grease.  To  get  the  lip  of  the  top  of  the  stationary  section  of  1 V4- 
inch  pipe  to  project  above  the  flange  plate^  it  will  be  necessary  to  have  a plumbing  shop  cut  a 
slightly  deeper  thread  inside  the  flange  plate,  as  well  as  cutting  an  unusually  long  thread  on 
the  end  of  the  1 ’/4-inch  pipe.  It  is  relatively  easy  to  waterproof  this  assembiy  through  the  roof 
since  the  7 ’4-inch  pipe  is  stationary  at  all  times.  Be  sure  to  use  pipe  compound  on  alt  the  joints 
and  then  really  tighten  these  joints  with  a pair  of  pipe  wrenches. 


512  Rotary  Beams 


the  radio 


has  been  determined  previously,  and  the 
Antennascope  control  is  turned  for  a null 
reading  on  the  meter  of  the  Antennascope. 
The  impedance  presented  to  the  Antenna- 
scope by  the  matching  device  may  be 
read  directly  on  the  calibrated  dial  of 
the  Antennascope. 

3.  Adjustments  should  be  made  to  the 
matching  device  to  present  the  desired 
impedance  transformation  to  the  Antenna- 
scope. If  a folded  dipole  is  used  as  the 
driven  element,  the  transformation  ratio 
of  the  dipole  must  be  varied  as  explained 
previously  in  this  chapter  to  provide  a 
more  exact  match.  If  a T-match  or  gamma 
match  system  is  used,  the  length  of  the 
matching  rod  may  be  changed  to  effect 
a proper  match.  If  the  Antennascope  ohm- 
ic reading  is  lower  than  the  desired  read- 
ing, the  length  of  the  matching  rod  should 
be  increased.  If  the  Antennascope  read- 
ing is  higher  than  the  desired  reading, 
the  length  of  the  matching  tod  should  be 
decreased.  After  each  change  in  length 
of  the  matching  tod,  the  series  capacitor 
in  the  matching  system  should  be  re- 
tesonated  for  best  null  on  the  meter  of 
the  Antennascope. 

Raising  and  A practical  problem  always  pres- 
Lowering  ent  when  tuning  up  and  matching 
the  Array  an  array  is  the  physical  location 
of  the  structure.  If  the  array  is 
atop  the  mast  it  is  inaccessible  for  adjustment, 
and  if  it  is  located  on  stepladders  where  it  can 
be  adjusted  easily  it  cannot  be  rotated.  One 
encouraging  factor  in  this  situation  is  the  fact 
that  experience  has  shown  that  if  the  array  is 
placed  8 or  10  feet  above  ground  on  some  step- 
ladders  for  the  preliminary  tuning  process,  the 
raising  of  the  system  to  its  full  height  will  not 
produce  a serious  change  in  the  adjustments. 
So  it  is  usually  possible  to  make  preliminary 
adjustments  with  the  system  located  slightly 
greater  than  head  height  above  ground,  and 
then  to  raise  the  antenna  to  a position  where 
it  may  be  rotated  for  final  adjustments.  If  the 
position  of  the  sliding  sections  as  determined 
near  the  ground  is  marked  so  that  the  adjust- 
ments will  not  be  lost,  the  array  may  be  raised 
to  rotatable  height  and  the  fastening  clamps 
left  loose  enough  so  that  the  elements  may  be 
slid  in  by  means  of  a long  bamboo  pole.  After 
a series  of  trials  a satisfactory  set  of  lengths 
can  be  obtained.  But  the  end  results  usually 
come  so  close  to  the  figures  given  in  figure  5 
that  a subsequent  array  is  usually  cut  to  the 
dimensions  given  and  installed  as-is. 

The  matching  process  does  not  requite  ro- 
tation, but  it  does  requite  that  the  antenna 
proper  be  located  at  as  nearly  its  normal  opet- 


Figupe  20 

HEAVY  DUTY  ROTATOR  SUITABLE 
FOR  AMATEUR  BEAMS 
The  new  Cornell. Dubilier  type  HAM.1 
rotor  has  extra  heavy  motor  and  gear. 
ing  system  to  withstand  weight  and 
Inertia  of  amateur  array  under  the 
buffeting  of  heavy  winds.  Steel  spur 
gears  and  rotor  lock  prevent  "pin. 
wheeling”  of  antenna. 


HANDBOOK 


Antenna  Control  Systems  513 


CONTROL  BOX 


ANTENNA  ROTATOR 


DIRECTION  INDICATOR 


Figure  21 

SCHEMATIC  OF  A COMPLETE  ANTENHA  CONTROL  SYSTEM 


ating  position  as  possible.  However,  on  a par- 
ticular installation  the  positions  of  the  current 
minimums  on  the  transmission  line  near  the 
transmitter  may  be  checked  with  the  array  in 
the  air,  and  then  the  array  may  be  lowered  to 
ascertain  whether  ot  not  the  positions  of  these 
points  have  moved.  If  they  have  not,  and  in 
most  cases  if  the  feeder  line  is  strung  out  back 
and  forth  well  above  ground  as  the  antenna  is 
lowered  they  will  not  change,  the  positions  of 
the  last  few  toward  the  antenna  itself  may  be 
determined.  Then  the  calculation  of  the  match- 
ing quartet-wave  section  may  be  made,  the  sec- 
tion installed,  the  standing-wave  ratio  again 
checked,  and  the  antenna  re-installed  in  its 
final  location. 


25-9  Antenna  Rotation  Systems 

Structures  for  the  rotation  of  antenna  arrays 
may  be  divided  into  two  general  classes:  the 
rotating  mast  and  the  rotating  platform.  The 
rotating  mast  is  especially  suitable  where  the 
transmitting  equipment  is  installed  in  the  gar- 
age or  some  structure  away  from  the  main 
house.  Such  an  installation  is  shown  in  figure 
19.  A very  satisfactory  rotation  mechanism  is 
obtained  by  the  use  of  a large  steering  wheel 


located  on  the  bottom  pipe  of  the  rotating  mast, 
with  the  thrust  bearing  for  the  structure  lo- 
cated above  the  roof. 

If  the  rotating  mast  is  located  a distance 
from  the  operating  position,  a system  of  pul- 
leys and  drive  rope  may  be  used  to  turn  the  an- 
tenna, or  a slow  speed  electric  motor  may  be 
employed. 

The  rotating  platform  system  is  best  if  a 
tower  or  telephone  pole  is  to  be  used  for  an- 
tenna support.  A number  of  excellent  rotating 
platform  devices  are  available  on  the  market 
for  varying  prices.  The  larger  and  more  expen- 
sive rotating  devices  are  suitable  for  the  to‘ta- 
of  a rather  sizeable  array  for  the  14-Mc.  band 
while  the  smaller  structures,  such  as  those 
designed  for  rotating  a TV  antenna  are  design- 
ed for  less  load  and  should  be  used  only  with 
a 28-Mc.  ot  50-Mc.  array.  Most  common  prac- 
tice is  to  install  the  rotating  device  atop  a 
platform  built  at  the  top  of  a telephone  pole 
or  on  the  top  of  a lattice  mast  of  sizeable  cross 
section  so  that  the  mast  will  be  self-support- 
ing and  capable  of  withstanding  the  torque  im- 
posed upon  it  by  the  rotating  platform. 

A heavy  duty  TV  rotator  may  be  employed 
for  rotation  of  6 and  10-meter  arrays.  Fifteen 
and  twenty  meter  arrays  should  use  rotators 
designed  for  amateur  use  such  as  the  Cornell- 
Dubilier  HAM-1  unit  shown  in  figure  20. 


514  Rotary  Beams 


25-10  Indication  of  Direction 


The  most  satisfactory  method  for  indicating 
the  direction  of  transmission  of  a rotatable 
array  is  that  which  uses  Selsyns  or  synchros 
for  the  transmission  of  the  data  from  the  ro- 
tating structure  to  the  indicating  pointer  at  the 
operating  position.  A number  of  synchros  and 
Selsyns  of  various  types  ate  available  on  the 
surplus  market.  Some  of  them  are  designed  for 
operation  on  115  volts  at  60  cycles,  some  are 
designed  for  operation  on  60  cycles  but  at  a 
lowered  voltage,  and  some  ate  designed  for 
operation  from  400-cycle  or  800-cycle  energy. 
This  latter  type  of  high-frequency  synchro  is 
the  most  generally  available  type,  and  the 
high-frequency  units  are  smaller  and  lighter 
than  the  60-cycle  units.  Since  the  indicating 
synchro  must  delivet  an  almost  negligible 
amount  of  power  to  the  pointer  which  it  drives, 
the  high-frequency  types  will  operate  quite 
satisfactorily  from  60-cycle  power  if  the  volt- 
age on  them  is  reduced  to  somewhere  between 
6.3  and  20  volts.  In  the  case  of  many  of  the 
units  available,  a connection  sheet  is  provided 
along  with  a recommendation  in  regard  to  the 
operating  voltage  when  they  are  run  on  60  cy- 
cles. In  any  event  the  operating  voltage  should 
be  held  as  low  as  it  may  be  and  still  give  sat- 
isfactory transmission  of  data  from  the  anten- 
na to  the  operating  position.  Cettainly  it  should 
not  be  necessary  to  run  such  a voltage  on  the 
units  that  they  become  overheated. 

A suitable  Selsyn  indicating  system  is  shown 
in  figure  21. 

Systems  using  a potentiometer  capable  of 
continuous  rotation  and  a milliammeter,  along 
with  a battery  or  other  source  of  direct  current, 
may  also  be  used  for  the  indication  of  direc- 
tion. A commercially-available  potentiometer 
(Ohmite  RB-2)  may  be  used  in  conjunction 
with  a 0-1  d-c  milliammeter  having  a hand- 
calibrated  scale  for  direction  indication. 

25-11  “Three-Bond”  Beams 

A popular  form  of  beam  antenna  introduced 
during  the  past  few  years  is  the  so-called 
three-band  beam.  An  array  of  this  type  is  de- 
signed to  operate  on  three  adjacent  amateur 
bands,  such  as  the  ten,  fifteen,  and  twenty 
meter  group.  The  principle  of  operation  of 
this  form  of  antenna  is  to  employ  parallel 
tuned  circuits  placed  at  critical  positions  in 
the  elements  of  the  beam  which  serve  to  elec- 
trically connect  and  disconnect  the  outer  sec- 
tions of  the  elements  as  the  frequency  of  ex- 
citation of  the  antenna  is  changed.  A typical 
“three-band”  element  is  shown  in  figure  22. 
At  the  lowest  operating  frequency,  the  tuned 


traps  exert  a minimum  influence  upon  the  ele- 
ment and  it  resonates  at  a frequency  determined 
by  the  electrical  length  of  the  configuration, 
plus  a slight  degree  of  loading  contributed  by 
the  traps.  At  some  higher  frequency  (generally 
about  1.5  times  the  lowest  operating  frequency) 
the  outer  set  of  traps  are  in  a parallel  resonant 
condition,  placing  a high  impedance  between 
the  element  and  the  tips  beyond  the  traps. 
Thus,  the  element  resonates  at  a frequency 
1.5  times  higher  than  that  determined  by  the 
overall  length  of  the  element.  As  the  frequency 
of  operation  is  raised  to  approximately  2.0 
times  the  lowest  operating  frequency,  the  inner 
set  of  traps  become  resonant,  effectively  dis- 
connecting a larger  portion  of  the  element  from 
the  driven  section.  The  length  of  the  center 
section  is  resonant  at  the  highest  frequency 
of  operation.  The  center  section,  plus  the  two 
adjacent  inner  sections  are  resonant  at  the 
intermediate  frequency  of  operation,  and  the 
complete  element  is  resonant  at  the  lowest 
frequency  of  operation. 

The  efficiency  of  such  a system  is  deter- 
mined by  the  accuracy  of  tuning  of  both  the 
element  sections  and  the  isolating  traps.  In 
addition  the  combined  dielectric  losses  of  the 
traps  affect  the  overall  antenna  efficiency. 
As  with  all  multi-purpose  devices,  some  com- 
promise between  operating  convenience  and 
efficiency  must  be  made  with  antennas  designed 
to  operate  over  more  than  one  narrow  band  of 
frequencies.  It  is  a tribute  to  the  designers  of 
the  better  multi -band  beams  that  they  perform 
as  well  as  they  do,  taking  into  account  the 
theoretical  difficulties  that  must  be  overcome. 


isolatinc  traps 


Figure  22 

TRAP-TYPE  "THREE  BAND- 
ELEMENT 


isolating  traps  permit  dipole  to  be 
self'-resonant  at  three  widely  different 
frequencies. 


CHAPTER  TWENTY-SIX 


Mobile  Equipment 
Design  and  Installation 


Mobile  operation  is  permitted  on  all  amateur 
bands.  Tremendous  impetus  to  this  phase  of 
the  hobby  was  given  by  the  suitable  design  of 
compact  mobile  equipment.  Complete  mobile 
installations  may  be  purchased  as  packaged 
units,  or  the  whole  mobile  station  may  be  home 
built,  according  to  the  whim  of  the  operator. 

The  problems  involved  in  achieving  a satis- 
factory two-way  installation  vary  somewhat 
with  the  band,  but  many  of  the  problems  are 
common  to  all  bands.  For  instance,  ignition 
noise  is  more  troublesome  on  10  meters  than 
on  75  meters,  but  on  the  other  hand  an  effi- 
cient antenna  system  is  much  more  easily  ac- 
complished on  10  meters  than  on  75  meters. 
Also,  obtaining  a worthwhile  amount  of  trans- 
mitter output  without  excessive  battery  drain 
is  a problem  on  all  bands. 


26-1  Mobile  Reception 

When  a broadcast  receiver  is  in  the  car,  the 
most  practical  receiving  arrangement  involves 
a converter  feeding  into  the  auto  set.  The  ad- 
vantages of  good  selectivity  with  good  image 

rejection  obtainable  from  a double  conversion 

superheterodyne  are  achieved  in  most  cases 
without  excessive  "birdie”  troubles,  a com- 


mon difficulty  with  a double  conversion  super- 
heterodyne constructed  as  an  integral  receiver 
in  one  cabinet.  However,  it  is  important  that 
the  b-c  receiver  employ  an  r-f  stage  in  order  to 
provide  adequate  isolation  between  the  con- 
verter and  the  high  frequency  oscillator  in  the 
b-c  receiver.  The  r-f  stage  also  is  desirable 
from  the  standpoint  of  image  rejection  if  the 
converter  does  not  employ  a tuned  output  cir- 
cuit (tuned  to  the  frequency  of  the  auto  set, 
usually  about  1500  kc.).  A few  of  the  late 
model  auto  receivers,  even  in  the  better  makes, 
do  not  employ  an  t-f  stage. 

The  usual  procedure  is  to  obtain  converter 
plate  voltage  from  the  auto  receiver.  Experi- 
ence has  shown  that  if  the  converter  does  not 
draw  more  than  about  15  or  at  most  20  ma.  tot- 
al plate  current  no  damage  to  the  auto  set  or 
loss  in  performance  will  occur  other  than  a 
slight  reduction  in  vibrator  life.  The  converter 
drain  can  be  minimized  by  avoiding  a voltage 
regulator  tube  on  the  converter  h-f  oscillator. 
On  10  meters  and  lower  frequencies  it  is  pos- 
sible to  design  an  oscillator  with  sufficient 
stability  so  that  no  voltage  regulator  is  re- 
quired in  the  converter. 

With  some  cats  satisfactory  75-meter  opera- 
tion can  be  obtained  without  a noise  clipper 
if  resistor  type  spark  plugs  (such  as  those 
made  by  Autolite)  ate  employed.  However,  a 
noise  clipper  is  helpful  if  not  absolutely  neces- 


515 


516  Mobile  Equipment 


THE  RADIO 


sary,  and  it  is  recommended  that  a noise  clip- 
per be  installed  without  confirming  the  neces- 
sity therefor.  It  has  been  found  that  quiet  re- 
ception sometimes  may  be  obtained  on  75  me- 
ters simply  by  the  use  of  resistor  type  plugs, 
but  after  a few  thousand  miles  these  plugs 
often  become  less  effective  and  no  longer  do 
a fully  adequate  job.  Also,  a noise  clipper  in- 
sures against  ignition  noise  from  passing 
trucks  and  "un-suppressed”  cars.  On  10  me- 
ters a noise  clipper  is  a “must”  in  any  case. 

Modifying  the  There  are  certain  things  that 
Auto  Receiver  should  be  done  to  the  auto  set 
when  it  is  to  be  used  with  a 
converter,  and  they  might  as  well  be  done  all 
at  the  same  time,  because  "dropping”  an  auto 
receiver  and  getting  into  the  chassis  to  work 
on  it  takes  quite  a little  time. 

First,  however,  check  the  circuit  of  the  auto 
receiver  to  see  whether  it  is  one  of  the  few 
receivers  which  employ  circuits  which  com- 
plicate connection  of  a noise  clipper  or  a con- 
verter. If  the  receiver  is  yet  to  be  purchased, 
it  is  well  to  investigate  these  points  ahead  of 
time. 

If  the  receiver  uses  a negative  B resistor 
strip  for  bias  (as  evidenced  by  the  cathode  of 
the  audio  output  stage  being  grounded),  then 
the  additional  plate  current  drain  of  the  con- 
verter will  upset  the  bias  voltages  on  the  var- 
ious stages  and  probably  cause  trouble.  Be- 
cause the  converter  is  not  on  all  the  time,  it 
is  not  practical  simply  to  alter  the  resistance 
of  the  bias  strip,  and  major  modification  of  the 
receiver  probably  will  be  required. 

The  best  type  of  receiver  for  attachment  of 
a converter  and  noise  clipper  uses  an  r-f  stage; 
permeability  tuning;  single  unit  construction 
(except  possibly  for  the  speaker);  push  button 
tuning  rather  than  a tuning  motor;  a high  vacu- 
um rectifier  such  as  a 6X4  (rather  than  an  0Z4 
or  a synchronous  rectifier);  a 6SQ7  (or  minia- 
ture or  Loctal  equivalent)  with  grounded  cath- 
ode as  second  detector,  first  audio,  and  a.v.c.; 
power  supply  negative  grounded  directly  (no 
common  bias  strip);  a PM  speaker  (to  minimize 
battery  drain);  and  an  internal  r-f  gain  control 
(indicating  plenty  of  built-in  reserve  gain 
which  may  be  called  upon  if  necessary).  Many 
current  model  auto  radios  have  all  of  the  fore- 
going features,  and  numerous  models  have  most 
of  them,  something  to  keep  in  mind  if  the  set 
is  yet  to  be  purchased. 

Noise  Limiters  A noise  limiter  either  may  be 
built  into  the  set  or  purchased 
as  a commercially  manufactured  unit  for  "out- 
board” connection  via  shielded  wires.  If  the 
receiver  employs  a 6SQ7  (or  Loctal  or  minia- 
ture equivalent)  in  a conventional  circuit,  it  is 
a simple  matter  to  build  in  a noise  clipper  by 


SWITCH  , SW 
(iN-OUT^ 


Figure  1 

SERIES-GATE  NOISE  LIMITER  FOR  AUTO 
RECEIVER 

Auto  receivers  using  a 6SQ7,  7B6,  7X7,  or 
6AT6  as  second  detector  and  o.v.c.  con  be 
converted  to  the  above  circuit  with  but  few 
wiring  changes.  The  circuit  has  the  advan- 
tage of  not  requiring  an  additional  tube  sock- 
et for  the  limiter  diode. 


substituting  a 6S8  octal,  7X7  Loctal,  or  a 6T8 
9-pin  miniature  as  shown  in  figure  1.  When 
substituting  a 6T8  for  a 6AT6  or  similar  7-pin 
miniature,  the  socket  must  be  changed  to  a 
9-pin  miniature  type.  This  requires  reaming 
the  socket  hole  slightly. 

If  the  receiver  employs  cathode  bias  on  the 
6SQ7  (or  equivalent),  and  perhaps  delayed 
a.v.c.,  the  circuit  usually  can  be  changed  to 
the  grounded-cathode  circuit  of  figure  1 with- 
out encountering  trouble. 

Some  receivers  take  the  t-f  excitation  for 
the  a-v-c  diode  from  the  plate  of  the  i-f  stage. 
In  this  case,  leave  the  a.v.c.  alone  and  ignore 
the  a-v-c  buss  connection  shown  in  figure  1 
(eliminating  the  1-megohm  decoupling  resistor). 
If  the  set  uses  a separate  a-v-c  diode  which 
receives  r-f  excitation  via  a small  capacitor 
connected  to  the  detector  diode,  then  simply 
change  the  circuit  to  correspond  to  figure  1. 

In  case  anyone  might  be  considering  the  use 
of  a crystal  diode  as  a noise  limiter  in  con- 
junction with  the  tube  already  in  the  set,  it 
might  be  well  to  point  out  that  crystal  diodes 
perform  quite  poorly  in  series-gate  noise  clip- 
pers of  the  type  shown. 

It  will  be  observed  that  no  tone  control  is 


HANDBOOK 


Mobile  Receivers  517 


shown.  Multi-position  tone  controls  tied  in 
with  the  second  detector  circuit  often  permit 
excessive  "leak  through.”  Hence  it  is  recom- 
mended that  the  tone  control  components  be 
completely  removed  unless  they  are  confined 
to  the  grid  of  the  a-f  output  stage.  If  removed, 
the  highs  can  be  attenuated  any  desired  amount 
by  connecting  a mica  capacitor  from  plate  to 
screen  on  the  output  stage.  Ordinarily  from 
.005  to  .01  pfd.  will  provide  a good  compro- 
mise between  fidelity  and  reduction  of  back- 
ground hiss  on  weak  signals. 

Usually  the  switch  SW  will  have  to  be 
mounted  some  distance  from  the  noise  limiter 
components.  If  the  leads  to  the  switch  ate  over 
af^roximately  IM  inches  long,  a piece  of 
shield  braid  should  be  slipped  over  them  and 
grounded.  The  same  applies  to  the  "hot”  leads 
to  the  volume  control  if  not  already  shielded. 
Closing  the  switch  disables  the  limiter.  This 
may  be  desirable  for  reducing  distortion  on 
broadcast  reception  or  when  checking  the  in- 
tensity of  ignition  noise  to  determine  the  ef- 
fectiveness of  suppression  measures  taken  on 
the  car.  The  switch  also  permits  one  to  check 
the  effectiveness  of  the  noise  clipper. 

The  22,000-ohm  decoupling  resistor  at  the 
bottom  end  of  the  i-f  transformer  secondary  is 
not  critical,  and  if  some  other  value  already 
is  incorporated  inside  the  shield  can  it  may  be 
left  alone  so  long  as  it  is  not  over  47,000 
ohms,  a common  value.  Higher  values  must  be 
replaced  with  a lower  value  even  if  it  requires 
a can  opener,  because  anything  over  47,000 
ohms  will  result  in  excessive  loss  in  gain. 
There  is  some  loss  in  a-f  gain  inherent  in  this 
type  of  limiter  anyhow  (slightly  over  6 db), 
and  it  is  important  to  minimize  any  additional 
loss. 

It  is  important  that  the  total  amount  of  ca- 
pacitance in  the  RC  decoupling  (t-f)  filter  not 
exceed  about  100  ^^fd.  With  a value  much 
greater  than  this  "pulse  stretching”  will  occur 
and  the  effectiveness  of  the  noise  clipper  will 
be  reduced.  Excessive  capacitance  will  reduce 
the  amplitude  and  increase  the  duration  of  the 
ignition  pulses  before  they  teach  the  clipper. 
The  reduction  in  pulse  amplitude  accomplices 
no  good  since  the  pulses  are  fed  to  the  clipper 
anyhow,  but  the  greater  duration  of  the  length- 
ened pulses  increases  the  audibility  and  the 
blanking  interval  associated  with  each  pulse. 
If  a shielded  wire  to  an  external  clipper  is  em- 
ployed, the  r-f  by-pass  on  the  "low”  side  of 
the  RC  filter  may  be  eliminated  since  the  ca- 
pacitance of  a few  feet  of  shielded  wire  will 
accomplish  the  same  result  as  the  by-pass 
capacitor. 

The  switch  SW  is  connected  in  such  a man- 
ner that  there  is  practically  no  change  in  gain 
with  the  limiter  in  or  out.  If  the  auto  set  does 
not  have  any  reserve  gain  and  more  gain  is 


needed  on  weak  broadcast  signals,  the  switch 
can  be  connected  from  the  hot  side  of  the  vol- 
ume control  to  the  junction  of  the  22,000, 
270,000  and  1 megohm  resistors  instead  of  as 
shown.  This  will  provide  approximately  6 db 
mote  gain  when  the  clipper  is  switched  out. 

Many  late  model  receivers  are  provided  with 
an  internal  t-f  gain  control  in  the  cathode  of 
the  r-f  and/or  i-f  stage.  This  control  should 
be  advanced  full  on  to  provide  better  noise 
limiter  action  and  make  up  for  the  loss  in  audio 
gain  introduced  by  the  noise  clipper. 

Installation  of  the  noise  clipper  often  de- 
tunes the  secondary  of  the  last  i-f  transformer. 
This  should  be  repeaked  before  the  set  is  per- 
manently replaced  in  the  cat  unless  the  trim- 
mer is  accessible  with  the  set  mounted  in 
place. 

Additional  clipper  circuits  will  be  found  in 
the  receiver  chapter  of  this  Handbook. 

Selectivity  While  not  of  serious  concern  on 
10  meters,  the  lack  of  selectivity 
exhibited  by  a typical  auto  receiver  will  result 
in  QRM  difficulty  on  20  and  75  meters.  A typi- 
cal auto  set  has  only  two  i-f  transformers  of 
relatively  low-Q  design,  and  the  second  one 
is  loaded  by  the  diode  detector.  The  skirt  se- 
lectivity often  is  so  poor  that  a strong  local 
will  depress  the  a.v.c.  when  listening  to  a 
weak  station  as  much  as  15  kc.  different  in 
frequency. 

One  solution  is  to  add  an  outboard  i-f  stage 
employing  two  good  quality  double-tuned  trans- 
formers (not  the  midget  variety)  connected 
"back-to-back”  through  a small  coupling  ca- 
pacitance. The  amplifier  tube  (such  as  a 6BA6) 
should  be  biased  to  the  point  where  the  gain 
of  the  outboard  unit  is  relatively  small  (1  or 
2),  assuming  that  the  receiver  already  has  ade- 
quate gain.  If  additional  gain  is  needed,  it  may 
be  provided  by  the  outboard  unit.  Low-capaci- 
tance shielded  cable  should  be  used  to  couple 
into  and  out  of  the  outboard  unit,  and  the  unit 
itself  should  be  thoroughly  shielded. 

Such  an  outboard  unit  will  sharpen  the  nose 
selectivity  slightly  and  the  skirt  selectivity 
greatly.  Operation  then  will  be  comparable  to 
a home-station  communications  receiver, 
though  selectivity  will  not  be  as  good  as  a 
receiver  employing  a 50-kc.  or  85-kc.  "Q5’er.” 

Obtaining  Power  While  the  set  is  on  the 

for  the  Converter  bench  for  installation  of 

the  noise  clipper,  provi- 
sion should  be  made  for  obtaining  filament  and 
plate  voltage  for  the  converter,  and  for  the  ex- 
citer and  speech  amplifier  of  the  transmitter, 
if  such  an  arrangement  is  to  be  used.  To  per- 
mit removal  of  either  the  converter  or  the  auto 
set  from  the  cat  without  removing  the  other,  a 
connector  should  be  provided.  The  best  method 


518  Mobile  Equipment 


THE  RADIO 


(OPTtONAu) 


Figure  2 

USING  THE  RECEIVER  PLATE  SUPPLY 
FOR  THE  TRANSMITTER 

This  circuit  silences  the  receiver  on  trans~ 
mit,  and  In  addition  makes  It  possible  to  use 
the  receiver  plate  supply  for  feeding  the  ex- 
alter  and  speech  amplifier  stages  In  the 
transmitter. 


is  to  mount  a small  receptacle  on  the  receiver 
cabinet  or  chassis,  making  connection  via  a 
matching  plug.  An  Amphenol  type  77-26  recep- 
tacle is  compact  enough  to  fit  in  a very  small 
space  and  allows  four  connections  (including 
ground  for  the  shield  braid).  The  matching  plug 
is  a type  70-26. 

To  avoid  the  possibility  of  vibrator  hash 
being  fed  into  the  converter  via  the  heater  and 
plate  voltage  supply  leads,  it  is  important  that 
the  heater  and  plate  voltages  be  taken  from 
points  well  removed  from  the  power  supply  por- 
tion of  the  auto  receiver.  If  a single-ended 
audio  output  stage  is  employed,  a safe  place 
to  obtain  these  voltages  is  at  this  tube  socket, 
the  high  voltage  for  the  converter  being  taken 
from  the  screen.  In  the  case  of  a push-pull  out- 
put stage,  however,  the  screens  sometimes  are 
fed  from  the  input  side  of  the  power  supply 
filter.  The  ripple  at  this  point,  while  suffi- 
ciently low  for  a push-pull  audio  output  stage, 
is  not  adequate  for  a converter  without  addi- 
tional filtering.  If  the  schematic  shows  that 


the  screens  of  a push-pull  stage  are  connected 
to  the  input  side  instead  of  the  output  side  of 
the  power  supply  filter  (usually  two  electroly- 
tics  straddling  a resistor  in  an  R-C  filter),  then 
follow  the  output  of  the  filter  over  into  the  r-f 
portion  of  the  set  and  pick  it  up  there  at  a con- 
venient point,  before  it  goes  through  any  addi- 
tional series  dropping  or  isolating  resistors, 
as  shown  in  figure  2. 

The  voltage  at  the  output  of  the  filter  usual- 
ly runs  from  200  to  250  volts  with  typical  con- 
verter drain  and  the  motor  not  running.  This 
will  increase  perhaps  10  per  cent  when  the 
generator  is  charging.  The  converter  drain  will 
drop  the  B voltage  slightly  at  the  output  of  the 
filter,  perhaps  15  to  25  volts,  but  this  reduc- 
tion is  not  enough  to  have  a noticeable  effect 
upon  the  operation  of  the  receiver.  If  the  B 
voltage  is  higher  than  desirable  or  necessary 
for  proper  operation  of  the  converter,  a 2-watt 
carbon  resistor  of  suitable  resistance  should 
be  inserted  in  series  with  the  plate  voltage 
lead  to  the  power  receptacle.  Usually  some- 
thing between  2200  and  4700  ohms  will  be 
found  about  right. 

Receiver  Disabling  When  the  battery  drain  is 
on  Transmit  high  on  transmit,  as  is  the 

case  when  a PE-103A  dy 
namotor  is  run  at  maximum  rating  and  other 
drains  such  as  the  transmitter  heaters  and  auto 
headlights  must  be  considered,  it  is  desirable 
to  disable  the  vibrator  power  supply  in  the  re- 
ceiver during  transmissions.  The  vibrator 
power  supply  usually  draws  several  amperes, 
and  as  the  receiver  must  be  disabled  in  some 
manner  anyhow  during  transmissions,  opening 
the  6-volt  supply  to  the  vibrator  serves  both 
purposes.  It  has  the  further  advantage  of  intro- 
ducing a slight  delay  in  the  receiver  recovery, 
due  to  the  inertia  of  the  power  supply  filter, 
thus  avoiding  the  possibility  of  a feedback 
"yoop”  when  switching  from  transmit  to  re- 
ceive. 

To  avoid  troubles  from  vibrator  hash,  it  is 
best  to  open  the  ground  lead  from  the  vibrator 
by  means  of  a midget  s.p.d.t.  6-volt  relay  and 
thus  isolate  the  vibrator  circuit  from  the  ex- 
ternal control  and  switching  circuit  wires.  The 
relay  is  hooked  up  as  shown  in  figure  3.'  Stand- 
ard 8-ampere  contacts  will  be  adequate  for 
this  application. 

The  relay  should  be  mounted  as  close  to 
the  vibrator  as  practicable.  Ground  one  of  the 
coil  terminals  and  run  a shielded  wire  from 
the  other  coil  terminal  to  one  of  the  power  re- 
ceptacle connections,  grounding  the  shield  at 
both  ends.  By-pass  each  end  of  this  wire  to 
ground  with  .01  fild.,  using  the  shortest  pos- 
sible leads.  A lead  is  run  from  the  correspond- 
ing terminal  on  the  mating  plug  to  the  control 
circuits,  to  be  discussed  later. 


HANDBOOK 


Mobile  Receiver  Installation  519 


Figure  3 

METHOD  OF  ELIMINATING  THE  BATTERY 
DRAIN  OF  THE  RECEIVER  VIBRATOR 
PACK  DURING  TRANSMISSION 

If  the  receiver  chassis  has  room  for  a mtrf* 
get  s.p,d,t,  reloYf  the  above  arrangement  not 
only  silences  the  receiver  an  transmit  but 
saves  several  amperes  battery  draim 


If  the  normally  open  contact  on  the  relay  is 
connected  to  the  hot  side  of  the  voice  coil 
winding  as  shown  in  figure  3 (assuming  one 
side  of  the  voice  coil  is  grounded  in  accord- 
ance with  usual  practice),  the  receiver  will 
be  killed  instantly  when  switching  from  re- 
ceive to  transmit,  in  spite  of  the  fact  that  the 
power  supply  filter  in  the  receiver  takes  a 
moment  to  discharge.  However,  if  a "slow 
start”  power  supply  (such  as  a dynamotor  or 
a vibrator  pack  with  a large  filter)  is  used  with 
the  transmitter,  shorting  the  voice  coil  prob- 
ably will  not  be  required. 

Using  the  Receiver  An  alternative  and  high- 
Plote  Supply  ly  recommended  proce- 

On  Transmit  dure  is  to  make  use  of 

the  receiver  B supply  on 
transmit,  instead  of  disabling  it.  One  disad- 
vantage of  the  popular  PE-103A  dynamotor  is 
the  fact  that  its  450-500  volt  output  is  too  high 
for  the  low  power  r-f  and  speech  stages  of  the 
transmitter.  Dropping  this  voltage  to  a more 
suitable  value  of  approximately  250  volts  by 
means  of  dropping  resistors  is  wasteful  of 
power,  besides  causing  the  plate  voltage  on 
the  oscillator  and  any  buffer  stages  to  vary 
widely  with  tuning.  By  means  of  a midget  6- 
volts.p.d.t.  relay  mounted  in  the  receiver,  con- 
nected as  shown  in  figure  2,  the  B supply  of 
the  auto  set  is  used  to  power  the  oscillator 
and  other  low  power  stages  (and  possibly 
screen  voltage  on  the  modulator).  On  transmit 
the  B voltage  is  removed  from  the  receiver 
and  converter,  automatically  silencing  the  re- 
ceiver. When  switching  to  receive  the  trans- 


mitter oscillator  is  killed  instantly,  thus  avoid- 
ing trouble  from  dynamotor  "carry  over.” 

The  efficiency  of  this  arrangement  is  good 
because  the  current  drain  on  the  main  high 
voltage  supply  for  the  modulated  amplifier  and 
modulator  plate (s)  is  reduced  by  the  amount 
of  current  borrowed  from  the  receiver.  At  least 
80  ma.  can  be  drawn  from  practically  all  auto 
sets,  at  least  for  a short  period,  without  dam- 
age. 

It  will  be  noted  that  with  the  arrangement  of 
figure  2,  plate  voltage  is  supplied  to  the  audio 
output  stage  at  all  times.  However,  when  the 
screen  voltage  is  removed,  the  plate  current 
drops  practically  to  zero. 

The  200-ohm  resistor  in  series  with  the  nor- 
mally open  contact  is  to  prevent  excessive 
sparking  when  the  contacts  close.  If  the  relay 
feeds  directly  into  a filter  choke  or  large  ca- 
pacitor there  will  be  excessive  sparking  at 
the  contacts.  Even  with  the  arrangement  shown, 
there  will  be  considerable  sparking  at  the  con- 
tacts; but  relay  contacts  can  stand  such  spark- 
ing quite  a while,  even  on  d.c.,  before  becom- 
ing worn  or  pitted  enough  to  require  attention. 
The  200-ohm  resistor  also  serves  to  increase 
the  effectiveness  of  the  .Ol-gfd.  r-f  by-pass 
capacitor. 

Auxiliary  Antenno  One  other  modification  of 
Trimmer  the  auto  receiver  which  may 

or  may  not  be  desirable  de- 
pending upon  the  circumstances  is  the  addi- 
tion of  an  auxiliary  antenna  trimmer  capacitor. 
If  the  converter  uses  an  untuned  output  circuit 
and  the  antenna  trimmer  on  the  auto  set  is 
peaked  with  the  converter  cut  in,  then  it  is 
quite  likely  that  the  trimmer  adjustment  will 
not  be  optimum  for  broadcast-band  reception 
when  the  converter  is  cut  out.  For  reception 
of  strong  broadcast  band  signals  this  usually 
will  not  be  serious,  but  where  reception  of 
weak  broadcast  signals  is  desired  the  loss  in 
gain  often  cannot  be  tolerated,  especially  in 
view  of  the  fact  that  the  additional  length  of 
antenna  cable  required  for  the  converter  in- 
stallation tends  to  reduce  the  strength  of 
broadcast  band  signals. 

If  the  converter  has  considerable  reserve 
gain,  it  may  be  practicable  to  peak  the  antenna 
trimmer  on  the  auto  set  for  broadcast-band  re- 
ception rather  than  resonating  it  to  the  con- 
verter output  circuit.  But  oftentimes  this  re- 
sults in  insufficient  converter  gain,  excessive 
image  troubles  from  loud  local  amateur  sta- 
tions, or  both. 

The  difficulty  can  be  circumvented  by  in- 
corporation of  an  auxiliary  antenna  trimmer 
connected  from  the  "hot”  antenna  lead  on  the 
auto  receiver  to  ground,  with  a switch  in 
series  for  cutting  it  in  or  out.  This  capacitor 
and  switch  can  be  connected  across  either  the 


52  0 Mobile  Equipment 


THE  RADIO 


converter  end  or  the  set  end  of  the  cable  be- 
tween the  converter  and  receiver.  This  auxil- 
iary trimmer  should  have  a range  of  about  3 
to  50  fifild.,  and  may  be  of  the  inexpensive 
compression  mica  type. 

With  the  trimmer  cut  out  and  the  converter 
turned  off  (by-passed  by  the  "in-out”  switch), 
peak  the  regular  antenna  trimmer  on  the  auto 
set  at  about  1400  kc.  Then  turn  on  the  convert- 
er, with  the  receiver  tuned  to  1500  kc.,  switch 
in  the  auxiliary  trimmer,  and  peak  this  trimmer 
for  maximum  background  noise.  The  auxiliary 
trimmer  then  can  be  left  switched  in  at  all 
times  except  when  receiving  very  weak  broad- 
cast band  signals. 

Some  auto  sets,  particularly  certain  General 
Motors  custom  receivers,  employ  a high-Q  high- 
impedance  input  circuit  which  is  very  critical 
as  to  antenna  capacitance.  Unless  the  shunt 
capacitance  of  the  antenna  (including  cable) 
approximates  that  of  the  anterma  installation 
for  which  the  set  was  designed,  the  antenna 
trimmer  on  the  auto  set  cannot  be  made  to  hit 
resonance  with  the  converter  cut  out.  This  is 
particularly  true  when  a long  antenna  cable  is 
used  to  reach  a whip  mounted  at  the  rear  of  the 
car.  Usually  the  condition  can  be  corrected  by 
unsoldering  the  internal  connection  to  the  an- 
tenna terminal  connector  on  the  auto  set  and 
inserting  in  series  a 100-ggfd.  mica  capacitor. 
Alternatively  an  adjustable  trimmer  covering 
at  least  50  to  150  pgfd.  may  be  substituted  for 
the  100-g^fd.  fixed  capacitor.  Then  the  adjust- 
ment of  this  trimmer  and  that  of  the  regular  an- 
tenna trimmer  can  be  juggled  back  and  forth 
until  a condition  is  achieved  where  the  input 
circuit  of  the  auto  set  is  resonant  with  the  con- 
verter either  in  or  out  of  the  circuit.  This  will 
provide  maximum  gain  and  image  rejection 
under  all  conditions  of  use. 

Reducing  Battery  When  the  receiving  installa- 
Drain  of  the  tion  is  used  frequently,  and 

Receiver  particularly  when  the  receiv- 

er is  used  with  the  car 
parked,  it  is  desirable  to  keep  the  battery 
drain  of  the  receiver-converter  installation  at 
an  absolute  minimum.  A substantial  reduction 
in  drain  can  be  made  in  many  receivers,  with- 
out appreciably  affecting  their  performance. 
The  saving  of  course  depends  upon  the  de- 
sign of  the  particular  receiver  and  upon  how 
much  trouble  and  expense  one  is  willing  to  go 
to.  Some  receivers  normally  draw  (without  the 
converter  connected)  as  much  as  10  amperes. 
In  many  cases  this  can  be  cut  to  about  5 am- 
peres by  incorporating  all  practicable  modi- 
fications. Each  of  the  following  modifications 
is  applicable  to  many  auto  receivers. 

If  the  receiver  uses  a speaker  with  a field 
coil,  replace  the  speaker  with  an  equivalent 
PM  type. 


Practically  all  0.3-ampere  r-f  and  a-f  volt- 
age amplifier  tubes  have  0.15-ampete  equiva- 
lents. In  many  cases  it  is  not  even  necessary 
to  change  the  socket  wiring.  However,  when 
substituting  i-f  tubes  it  is  recommended  that 
the  i-f  trimmer  adjustments  be  checked.  Gen- 
erally speaking  it  is  not  wise  to  attempt  to 
substitute  for  the  converter  tube  or  a-f  power 
output  tube. 

If  the  a-f  output  tube  employs  conventional 
cathode  bias,  substitute  a cathode  resistor  of 
twice  the  value  originally  employed,  or  add 
an  identical  resistor  in  series  with  the  one 
already  in  the  set.  This  will  reduce  the  B 
drain  of  the  receiver  appreciably  without  ser- 
iously reducing  the  maximum  undistorted  out- 
put. Because  the  vibrator  power  supply  is  much 
less  than  100  per  cent  efficient,  a saving  of 
one  watt  of  B drain  results  in  a saving  of  near- 
ly 2 watts  of  battery  drain.  This  also  mini- 
mizes the  overload  on  the  B supply  when  the 
converter  is  switched  in,  assuming  that  the 
converter  uses  B voltage  from  the  auto  set. 

If  the  receiver  uses  push-pull  output  and  if 
one  is  willing  to  accept  a slight  reduction  in 
the  maximum  volume  obtainable  without  dis- 
tortion, changing  over  to  a single  ended  stage 
is  simple  if  the  receiver  employs  conventional 
cathode  bias.  Just  pull  out  one  tube,  double 
the  value  of  cathode  bias  resistance,  and  add 
a 25-ftfd.  by-pass  capacitor  across  the  cathode 
resistor  if  not  already  by-passed.  In  some 
cases  it  may  be  possible  to  remove  a phase 
inverter  tube  along  with  one  of  the  a-f  output 
tubes. 

If  the  receiver  uses  a motor  driven  station 
selector  with  a control  tube  (d-c  amplifier), 
usually  the  tube  can  be  removed  without  up- 
setting the  operation  of  the  receiver.  One  then 
must  of  course  use  manual  tuning. 

While  the  changeover  is  somewhat  expen- 
sive, the  0.6  ampere  drawn  by  a 6X4  or  6X5 
rectifier  can  be  eliminated  by  substituting  six 
115-volt  r-m-s  50-ma.  selenium  rectifiers  (such 
as  Federal  type  402D3200).  Three  in  series 
are  substituted  for  each  half  of  the  full-wave 
rectifier  tube.  Be  sure  to  observe  the  correct 
polarity.  The  selenium  rectifiers  also  make  a 
good  substitution  for  an  0Z4  or  0Z4-GT  which 
is  causing  hash  difficulties  when  using  the 
converter. 

Offsetting  the  total  cost  of  nearly  $4.00  is 
the  fact  that  these  rectifiers  probably  will  last 
for  the  entire  life  of  the  auto  set.  Before  pur- 
chasing the  rectifiers,  make  sure  that  there  is 
room  available  for  mounting  them.  While  these 
units  are  small,  most  of  the  newer  auto  sets 
employ  very  compact  construction. 

Two-Meter  Reception  For  reception  on  the  144- 
Mc.  amateur  band,  and 
those  higher  in  frequency,  the  simple  converter- 


HANDBOOK 


One-Tube  Converter 


521 


auto-set  combination  has  not  proven  very  satis- 
factory. The  primary  reason  for  this  is  the  fact 
that  the  relatively  sharp  i-f  channel  of  the  auto 
set  imposes  too  severe  a limitation  on  the  sta- 
bility of  the  high-frequency  oscillator  in  the 
converter.  And  if  a crystal-controlled  beating 
oscillator  is  used  in  the  converter,  only  a por- 
tion of  the  band  may  be  coveted  by  tuning  the 
auto  set. 

The  most  satisfactory  arrangement  has  been 
found  to  consist  of  a separately  mounted  i.f., 
audio,  and  power  supply  system,  with  the  con- 
verter mounted  near  the  steering  column.  The 
i-f  system  should  have  a bandwidth  of  30  to 
100  kc.  and  may  have  a center  frequency  of 
10.7  Me.  if  standard  i-f  transformers  ate  to  be 
used.  The  control  head  may  include  the  144- 
Mc.  r-f,  mixer,  and  oscillator  sections,  and 
sometimes  the  first  i-f  stage.  Alternatively, 
the  control  head  may  include  only  the  h-f  os- 
cillator, with  a broadband  r-f  unit  included 
within  the  main  receiver  assembly  along  with 
the  i.f.  and  audio  system.  Commercially  manu- 
factured kits  and  complete  units  using  this 
general  lineup  are  available. 

An  alternative  arrangement  is  to  build  a 
converter,  10.7-Mc.  i-f  chatmel,  and  second 
detector  unit,  and  then  to  operate  this  unit  in 
conjunction  with  the  auto-set  power  supply, 
audio  system,  and  speaker.  Such  a system 
makes  economical  use  of  space  and  power 
drain,  and  can  be  switched  to  provide  normal 
broadcast-band  auto  reception  or  reception 
through  a converter  for  the  h-f  amateur  bands. 

A recent  development  has  been  the  VHP 
transceiver,  typified  by  the  Gonset  Communi- 
cator. Such  a unit  combines  a crystal  con- 
trolled transmitter  and  a tunable  VHP  receiver 
together  with  a common  audio  system  and 
power  supply.  The  complete  VHP  station  may 
be  packaged  in  a single  cabinet.  Various 
forms  of  VHP  transceivers  are  shown  in  the 
construction  chapters  of  this  Handbook. 


26-2  Mobile  Transmitters 

As  in  the  case  of  transmitters  for  fixed-sta- 
tion operation,  there  are  many  schools  of 
thought  as  to  the  type  of  transmitter  which  is 
most  suitable  for  mobile  operation.  One  school 
states  that  the  mobile  transmitter  should  have 
very  low  power  drain,  so  that  no  modification 
of  the  electrical  system  of  the  automobile  will 
be  required,  and  so  that  the  equipment  may  be 
Operated  wi±out  serious  regard  to  discharging 
the  battery  when  the  car  is  stopped,  or  over- 
loading the  generator  when  the  car  is  in  mo- 


tion. A total  transmitter  power  drain  of  about 
80  watts  from  the  car  battery  (6  volts  at  13 
amperes,  or  12  volts  at  7 amperes)  is  about 
the  maximum  that  can  be  allowed  under  these 
conditions.  For  maximum  power  efficiency  it 
is  recommended  that  a vibrator  type  of  supply 
be  used  as  opposed  to  a dynamotor  supply, 
since  the  conversion  efficiency  of  the  vibrator 
unit  is  high  compared  to  that  of  the  dynamotor. 

A second  school  of  thought  states  that  the 
mobile  transmitter  should  be  of  relatively  high 
power  to  overcome  the  poor  efficiency  of  the 
usual  mobile  whip  antenna.  In  this  case,  the 
mobile  power  should  be  drawn  from  a system 
that  is  independent  from  the  electrical  system 
of  the  automobile.  A belt  driven  high  voltage 
generator  is  often  coupled  to  the  automobile 
engine  in  this  type  of  installation. 

A variation  of  this  idea  is  to  employ 
a complete  secondary  power  system  in  the 
car  capable  of  providing  115  volts  a.c.  Shown 
in  figure  4 is  a Leece-Neville  three  phase 
alternator  mounted  atop  the  engine  block,  and 
driven  with  a fan  belt.  The  voltage  regulator 
and  selenium  rectifier  for  charging  the  car 
battery  from  the  a-c  system  replace  the  usual 
d-c  generator.  These  new  items  are  mounted 
in  the  front  of  the  car  radiator.  The  alternator 
provides  a balanced  delta  output  circuit  where- 
in the  line  voltage  is  equal  to  the  coil  volt- 
age, but  the  line  current  times  the  coil 

current.  The  coil  voltage  is  a nominal  6-volts, 
RMS  and  three  6.3  volt  25  ampere  filament 
transformers  may  be  connected  in  delta  on 
the  primary  and  secondary  windings  to  step 
the  6-volts  up  to  three-phase  115-volts.  If 
desired,  a special  115-volt,  3-phase  step-up 
transformer  may  be  wound  which  will  occupy 
less  space  than  the  three  filament  trans- 
formers. Since  the  ripple  frequency  of  a three 
phase  d-c  power  supply  will  be  quite  high,  a 
single  10  mfd  filter  capacitor  will  suffice. 


Figure  4 

LEECE-NEVILLE  3-PHASE 
ALTERNATOR  IS  ENGINE  DRIVEN 
BY  AUXILIARY  FAN  BELT. 


522  Mobile  Equipment 


THE  RADIO 


Figure  5 

A CENTER  LOADED  80-METER  WHIP 
USING  AIR  WOUND  COIL  MAY  BE  USED 
WITH  HIGH  POWERED  TRANSMITTERS 

An  anti~corona  loop  is  placed  at  the  top 
of  the  whip  to  reduce  loss  of  power  and 
burning  of  tip  of  antenna.  Number  of  forns 
in  coil  is  critical  and  adjustable,  high-Q 
coil  is  recommended.  Whip  may  be  used 
over  frequency  range  of  about  IS  kilo- 
cycles  without  retuning. 


26-3  Antennas  for 

Mobile  Work 

10-Meter  Mobile  The  most  popular  mobile  an- 
Antennos  teima  for  10-meter  operation 

is  a rear-mounted  whip  approx- 
imately 8 feet  long,  fed  with  coaxial  line.  This 
is  a highly  satisfactory  antenna,  but  a few 


5/16.WAVE  WHIP  RADIATOR  FOR  10 
METERS 

If  a whip  antenna  is  made  slightly  longer 
than  one~quarter  wove  it  acts  as  a slightly 
better  radiator  than  the  usual  quarter^wave 
whip,  and  it  can  provide  a better  match  to 
the  antenna  transmission  line  if  the  react- 
ance is  tuned  out  by  a series  capacitor  close 
to  the  base  of  the  antenna.  Capacitor  Cj  may 
be  a 100-p.iifd,  midget  variable. 


remarks  are  in  order  on  the  subject  of  feed 
and  coupling  systems. 

The  feed  point  resistance  of  a resonant  quar- 
tet-wave tear-mounted  whip  is  approximately 
20  to  25  ohms.  While  the  standing-wave  ratio 
when  using  50-ohm  coaxial  line  will  not  be 
much  greater  than  2 to  1,  it  is  nevertheless 
desirable  to  make  the  line  to  the  transmitter 
exactly  one  quarter  wavelength  long  electri- 
cally at  the  center  of  the  band.  This  procedure 
will  minimize  variations  in  loading  over  the 
band.  The  physical  length  of  RG-8/U  cable, 
from  antenna  base  to  antenna  coupling  coil, 
should  be  approximately  5 feet  3 inches.  The 
antenna  changeover  relay  preferably  should  be 
located  either  at  the  antenna  end  or  the  trans- 
mitter end  of  the  line,  but  if  it  is  mote  con- 
venient physically  the  line  may  be  broken  any- 
where for  insertion  of  the  relay. 

If  the  same  rear-mounted  whip  is  used  for 
broadcast-band  reception,  attenuation  of  broad- 
cast-band signals  by  the  high  shunt  capaci- 
tance of  the  low  impedance  feed  line  can  be 
reduced  by  locating  the  changeover  relay  tight 
at  the  antenna  lead  in,  and  by  running  95-ohm 
coax  (instead  of  50  or  75  ohm  coax)  from  the 
relay  to  the  converter.  Ordinarily  this  will  pro- 
duce negligible  effect  upon  the  operation  of 
the  converter,  but  usually  will  make  a worth- 
while improvement  in  the  strength  of  broadcast- 
band  signals. 


HANDBOOK 


Mobile  Antennas  523 


A more  effective  radiator  and  a better  line 
match  may  be  obtained  by  making  the  whip 
af^roximately  10  feet  long  and  feeding  it 
with  75-ohm  coax  (such  as  RG-ll/U)  via  a 
series  capacitor,  as  shown  in  figure  6.  The 
relay  and  series  capacitor  are  mounted  inside 
the  trunk,  as  close  to  the  antenna  feedthrough 
or  base-mount  insulator  as  possible.  The  10^- 
foot  length  applies  to  the  overall  length  from 
the  tip  of  the  whip  to  the  point  where  the  lead 
in  passes  through  the  car  body.  The  leads  in- 
side the  car  (connecting  the  coaxial  cable, 
relay,  series  capacitor  and  antenna  lead) 
should  be  as  short  as  possible.  The  outer  con- 
ductor of  both  coaxial  cables  should  be  ground- 
ed to  tbe  car  body  at  the  relay  end  with  short, 
heavy  conductors. 

A 100-/i/ifd.  midget  variable  capacitor  is 
suitable  for  C,.  The  optimum  setting  should 
be  determined  experimentally  at  the  center  of 
the  band.  This  setting  then  will  be  satisfac- 
tory over  the  whole  band. 

One  suitable  coupling  arrangement  for  either 
a ^-wave  or  5/16- wave  whip  on  10  meters  is 
to  use  a conventional  tank  circuit,  inductively 
coupled  to  a "variable  link”  coupling  loop 
which  feeds  the  coaxial  line.  Alternatively,  a 
pi-network  output  circuit  may  be  used.  If  the 
input  impedance  of  the  line  is  very  low  and 
the  tank  circuit  has  a low  C/L  ratio,  it  may 
be  necessary  to  resonate  the  coupling  loop 
with  series  capacitance  in  order  to  obtain  suf- 
ficient coupling.  This  condition  often  is  en- 
countered with  a )4-wave  whip  when  the  line 
length  approximates  an  electrical  half  wave- 
length. 

If  an  all-band  center-loaded  mobile  antenna 
is  used,  the  loading  coil  at  the  center  of  the 
antenna  may  be  shorted  out  for  operation  of 
the  antenna  on  the  10-meter  band.  The  usual 
type  of  center-loaded  mobile  antenna  will  be 
between  9 and  11  feet  long,  including  the  cen- 
ter-loading inductance  which  is  shorted  out. 
Hence  such  an  antenna  may  be  shortened  to 
an  electrical  quarter-wave  for  the  10-meter 
band  by  using  a series  capacitor  as  just  dis- 
cussed. Alternatively,  if  a pi-network  is  used 
in  the  plate  circuit  of  the  output  stage  of  the 
mobile  transmitter,  any  reactance  presented 
at  the  antenna  terminals  of  the  transmitter  by 
the  antenna  may  be  tuned  out  with  the  pi-net- 
work. 

The  All-Band  The  great  majority  of  mobile 
Center-Loaded  operation  on  the  i4-Mc.  band 
Mobile  Antenna  and  below  is  with  center 
loaded  whip  antennas.  These 
antennas  use  an  insulated  bumper  or  body 
mount,  with  provision  for  coaxial  feed  from 
the  base  of  the  antenna  to  the  transmitter,  as 
shown  in  figure  7. 

The  center-loaded  whip  antenna  must  be 


THE  CENTER-LOADED  WHIP  ANTENNA 
The  center-loaded  whip  anfenna,  when  pro- 
vided with  a tapped  loading  coil  or  a series 
of  coilsf  may  be  used  over  a wide  frequency 
range.  The  loading  coil  may  be  shorted  for 
use  of  the  antenna  on  the  I0~meter  band. 


tuned  to  obtain  optimum  operation  on  the  de- 
sired frequency  of  operation.  These  antennas 
will  operate  at  maximum  efficiency  over  a 
range  of  perhaps  20  kc.  on  the  75-meter  band, 
covering  a somewhat  wider  range  on  the  40- 
meter  band,  and  covering  the  whole  20-meter 
phone  band.  The  procedure  for  tuning  the  an- 
tennas is  discussed  in  the  instruction  sheet 
which  is  furnished  with  them,  but  basically 
the  procedure  is  as  follows: 

The  antenna  is  installed,  fully  assembled, 
with  a coaxial  lead  of  RG-58/U  from  the  base 
of  the  antenna  to  the  place  where  the  trans- 
mitter is  installed.  The  rear  deck  of  the  cat 
should  be  closed,  and  the  car  should  be  parked 
in  a location  as  clear  as  possible  of  trees, 
buildings,  and  overhead  power  lines.  Objects 
within  15  or  20  feet  of  the  antenna  can  exert 
a considerable  detuning  effect  on  the  antenna 
system  due  to  its  relatively  high  operating  Q. 
The  end  of  the  coaxial  cable  which  will  plug 
into  the  transmitter  is  terminated  in  a link  of 
3 or  4 turns  of  wire.  This  link  is  then  coupled 
to  a grid-dip  meter  and  the  resonant  frequency 
of  the  antenna  determined  by  noting  the  fre- 
quency at  which  the  grid  current  fluctuates. 
The  coils  furnished  with  the  antennas  normal- 
ly are  too  large  for  the  usual  operating  fre- 
quency, since  it  is  much  easier  to  remove 
turns  than  to  add  them.  Turns  then  are  removed, 
one  at  a time,  until  the  antenna  resonates  at 
the  desired  frequency.  If  too  many  turns  have 
been  removed,  a length  of  wire  may  be  spliced 

on  and  soldered.  Then,  with  a length  of  insu- 
lating tubing  slipped  over  the  soldered  joint, 
turns  may  be  added  to  lower  the  resonant  fre- 


524  Mobile  Equipment 


THE  RADIO 


Figure  8 

PI-NETWORK  ANTENNA  COUPLER 
The  pi-network  antenna  coupler  is  particu- 
larly satisfactory  for  mobile  work  since  the 
coupler  affords  some  degree  of  harmonic  re- 
ductiong  provides  a coupling  variation  to 
meet  varying  load  conditions  caused  by  fre- 
quency changes,  and  can  cancel  out  react- 
ance presented  to  the  transmitter  at  the  end 
of  the  antenna  transmission  line* 

For  use  of  the  coupler  on  the  3,9-Mc*  band 
Cj  should  have  a maximum  capacitance  of 
about  250  fMfifd*,  Lj  should  be  about  9 ml- 
crohenrys  (30  turns  J dia*  by  2 long),  and 
C2  may  include  a fixed  and  a variable  ele- 
ment with  maximum  capacitance  of  about  1400 
p-fJ-fd*  A 100-fip.fd,  variable  capacitor  will  be 
suitable  at  Cj  for  the  74-Mc.  and  28-Mc, 
bands,  with  a 350-p.fJ.fd.  variable  at  In- 
ductor Lj  should  have  an  Inductance  of  a- 
bout  2 m/cfohenrys  (H  turns  1"  dia,  by  J " 
long)  for  the  14-Me*  band,  and  about  0*8  mi- 
crohenry (6  turns  1 dia*  by  1*'  long)  for  the 
28-Mc*  bend* 


quency.  Or,  if  the  tapped  type  of  coil  is  used, 
taps  are  changed  until  the  proper  number  of 
turns  for  the  desired  operating  frequency  is 
found.  This  procedure  is  repeated  for  the  dif- 
ferent bands  of  operation. 


Feeding  the  After  much  experimenting  it 

Center-Loaded  has  been  found  that  the  most 
Antenna  satisfactory  method  for  feed- 

ing the  coaxial  line  to  the 
base  of  a center-loaded  antenna  is  with  the 
pi-network  coupler.  Figure  8 shows  the  basic 
arrangement,  with  recommended  circuit  con- 
stants. It  will  be  noted  that  relatively  large 
values  of  capacitance  are  required  for  all  bands 
of  operation,  with  values  which  seem  particu- 
larly large  for  the  75*meter  band.  But  refer- 
ence to  the  discussion  of  pi-network  tank  cir- 
cuits in  Chapter  13  will  show  that  the  val- 
ues suggested  are  normal  for  the  values  of  im- 
pedance, impedance  transformation,  and  oper- 
ating Q which  are  encountered  in  a mobile  in- 
stallation of  the  usual  type. 


26-4  Construction  and 

Installation  of  Mobile 
Equipment 

It  is  recommended  that  the  following  mea- 
sures be  taken  when  constructing  mobile  equip- 
ment, either  transmitting  or  receiving,  to  en- 
sure trouble-free  operation  over  long  periods; 

Use  only  a stiff,  heavy  chassis  unless  the 
chassis  is  quite  small. 

Use  lock  washers  or  lock  nuts  when  mount- 
ing components  by  means  of  screws. 

Use  stranded  hook-up  wire  except  where  r-f 
considerations  make  it  inadvisable  (such  as 
for  instance  the  plate  tank  circuit  leads  in  a 
v-h-f  amplifier).  Lace  and  tie  leads  wherever 
necessary  to  keep  them  from  vibrating  or  flop- 
ping around. 

Unless  provided  with  gear  drive,  tuning  ca- 
pacitors in  the  large  sizes  will  require  a rotor 
lock. 

Filamentary  (quick  heating)  tubes  should 
be  mounted  only  in  a vertical  position. 

The  larger  size  carbon  resistors  and  mica 
capacitors  should  not  be  supported  from  tube 
socket  pins,  particularly  from  miniature  sock- 
ets. Use  tie  points  and  keep  the  resistor  and 
capacitor  "pigtails”  short. 

Generally  speaking,  rubber  shock  mounts 
ate  unnecessary  or  even  undesirable  with  pas- 
senger cat  installations,  or  at  least  with  full 
size  passenger  cars.  The  springing  is  suffi- 
ciently "soft”  that  well  constructed  radio 
equipment  can  be  bolted  directly  to  the  vehicle 
without  damage  from  shock  or  vibration.  Un- 
less shock  mounting  is  properly  engineered 
as  to  the  stiffness  and  placement  of  the  shock 
mounts,  mechanical-resonance  "amplification” 
effects  may  actually  cause  the  equipment  to 
be  shaken  mote  than  if  the  equipment  were 
bolted  directly  to  the  vehicle. 

Surplus  military  equipment  provided  with 
shock  or  vibration  mounts  was  intended  for 
use  in  aircraft,  jeeps,  tanks,  gun-firing  Naval 
craft,  small  boats,  and  similar  vehicles  and 
craft  subject  to  severe  shock  and  vibration. 
Also,  the  shock  mounting  of  such  equipment 
is  very  carefully  engineered  in  order  to  avoid 
harmful  resonances. 

To  facilitate  servicing  of  mobile  equipment, 
all  interconnecting  cables  between  units 
should  be  provided  with  separable  connectors 
on  at  least  one  end. 


Control  Circuits  The  send-receive  control  cir- 
cuits of  a mobile  installation 
are  dictated  by  the  design  of  the  equipment, 
and  therefore  will  be  left  to  the  ingenuity  of 
the  reader.  However,  a few  generalizations 
and  suggestions  are  in  order. 


HANDBOOK 


Control  Circuits  525 


Do  not  attempt  to  control  too  many  relays, 
particularly  heavy  duty  relays  with  large  coils, 
by  means  of  an  ordinary  push-to-talk  switch 
on  a microphone.  These  contacts  are  not  de- 
signed for  heavy  work,  and  the  inductive  kick 
will  cause  more  sparking  than  the  contacts  on 
the  microphone  switch  ate  designed  to  handle. 
It  is  better  to  actuate  a single  relay  with  the 
push-to-talk  switch  and  then  control  all  other 
relays,  including  the  heavy  duty  contactor  for 
the  dynamotor  or  vibrator  pack,  with  this  relay. 

The  procedure  of  operating  only  one  relay 
direcdy  by  the  push-to-talk  switch,  with  all 
other  relays  being  controlled  by  this  control 
relay,  will  eliminate  the  often-encountered  dif- 
ficulty where  the  shutting  down  of  one  item  of 
equipment  will  close  relays  in  other  items  as 
a result  of  the  coils  of  relays  being  placed  in 
series  with  each  other  and  with  heater  circuits. 
A recommended  general  control  circuit,  where 
one  side  of  the  main  control  relay  is  connected 
to  the  hot  6-volt  circuit,  but  all  other  relays 
have  one  side  connected  to  ground,  is  illus- 
trated in  figure  9-  An  additional  advantage 
of  such  a circuit  is  that  only  one  control  wire 
need  be  run  to  the  coil  of  each  additional  re- 
lay, the  other  side  of  the  relay  coils  being 
grounded. 

The  heavy-duty  6-volt  solenoid-type  contac- 
tor relays  such  as  provided  on  the  PE-103A 
and  used  for  automobile  starter  relays  usually 
draw  from  1.5  to  2 amperes.  While  somewhat 
more  expensive,  heavy-duty  6-volt  relays  of 
conventional  design,  capable  of  breaking  30 
amperes  at  6 volts  d.c.,  ate  available  with 
coils  drawing  less  than  0.5  ampere. 

When  purchasing  relays  keep  in  mind  that 
the  current  rating  of  the  contacts  is  not  a fixed 
value,  but  depends  upon  (1)  the  voltage,  (2) 
whether  it  is  a.c.  or  d.c.,  and  (3)  whether  the 
circuit  is  purely  resistive  or  is  inductive.  If 
in  doubt,  refer  to  the  manufacturer’s  recom- 
mendations. Also  keep  in  mind  that  a dynamo- 
tor  presents  almost  a dead  short  until  the  ar- 
mature starts  turning,  and  the  starting  relay 
should  be  rated  at  considerably  more  than  the 
normal  dynamotor  current. 

Microphones  The  most  generally  used  micro- 
ond  Circuits  phone  for  mobile  work  is  the 
single-button  carbon.  With  a 
high-output-type  microphone  and  a high-ratio 
microphone  transformer,  it  is  possible  when 
"close  talking”  to  drive  even  a pair  of  push- 
pull  6l6’s  without  resorting  to  a speech  am- 
plifier. However,  there  is  a wide  difference  in 
the  output  of  the  various  type  single  button 
microphones,  and  a wide  difference  in  the  a- 
mount  of  step  up  obtained  with  different  type 
microphone  transformers.  So  at  least  one 
speech  stage  usually  is  desirable. 

One  of  the  most  satisfactory  single  button 


PUSH-TO-TALK  PUSH-TO-TALK 
SWITCH  ON  MIKE  RELAY 


ALTERNATE 

CONTROL 

SWITCH 


MAIN  POWER  RECEIVER 
RELAY  MUTING 
RELAY 


ANTENNA 

CHANGEOVER 

RELAY 


ANY 

OTHER 

RELAYS 


Figure  9 

RELAY  CONTROL  CIRCUIT 
Simplified  schematic  of  the  recommended 
relay  control  circuit  for  mobile  transmitters. 
The  relatively  small  push-to-talk  relay  Is 
controlled  by  the  button  on  the  microphone 
or  the  communications  switch.  Then  one  of 
the  contacts  on  this  relay  controls  the  other 
relays  of  the  transmitter;  one  side  of  the 
coil  of  all  the  additional  relays  controlled 
should  be  grounded. 


microphones  is  the  standard  Western  Electric 
type  F-1  unit  (or  Automatic  Electric  Co.  equi- 
valent). This  microphone  has  very  high  output 
when  operated  at  6 volts,  and  good  fidelity 
on  speech.  When  used  without  a speech  am- 
plifier stage  the  microphone  transformer  should 
have  a 50-ohm  primary  (rather  than  200  or  500 
ohms)  and  a secondary  of  at  least  150,000 
ohms  and  preferably  250,000  ohms. 

The  widely  available  surplus  type  T-17  mi- 
crophone has  higher  resistance  (200  to  500 
ohms)  and  lower  output,  and  usually  will  re- 
quire a stage  of  speech  amplification  except 
when  used  with  a very  low  power  modulator 
stage. 

Unless  an  F-1  unit  is  used  in  a standard 
housing,  making  contact  to  the  button  presents 
somewhat  of  a problem.  No  serious  damage  will 
result  from  soldering  to  the  button  if  the  con- 
nection is  made  to  one  edge  and  the  solder- 
ing is  done  very  rapidly  with  but  a small  a- 
mount  of  solder,  so  as  to  avoid  heating  the 
whole  button. 

A sound-powered  type  microphone  removed 
from  one  of  the  chest  sets  available  in  the 
surplus  market  will  deliver  almost  as  much 
voltage  to  the  grid  of  a modulator  stage  when 
used  with  a high-ratio  microphone  transformer 
as  will  an  F-1  unit,  and  has  the  advantage  of 
not  requiring  button  current  or  a **hash  filter.** 
This  is  simply  a dynamic  microphone  designed 
for  high  output  rather  than  maximum  fidelity. 

The  standardized  connections  for  a single- 

button  carbon  microphone  provided  with  push- 

to-talk  switch  are  shown  in  figure  10.  Prac- 
tically all  hand-held  military-type  single-button 


526  Mobile  Equipment 


THE  RADIO 




Xld 

PRESS-TO-TALK 
SWITCH 

Figure  10 

STANDARD  CONNECTIONS  FOR  THE 
PUSH-TO-TALK  SWITCH  ON  A HAND- 
HELD  SINGLE-BUTTON  CARBON 
MICROPHONE 


microphones  on  the  surplus  market  use  these 
connections. 

There  is  an  increasing  tendency  among  mo- 
bile operators  toward  the  use  of  microphones 
having  better  frequency  and  distortion  char- 
acteristics than  the  standard  single-button 
type.  The  high- impedance  dynamic  type  is 
probably  the  most  popular,  with  the  ceramic^ 
crystal  type  next  in  popularity.  The  conven- 
tional crystal  type  is  not  suitable  for  mobile 
use  since  the  crystal  unit  will  be  destroyed 
by  the  high  temperatures  which  can  be  reached 
in  a closed  car  parked  in  the  sun  in  the  sum- 
mer time. 

The  use  of  low-level  microphones  in  mobile 
service  requires  careful  attention  to  the  elimi- 
nation of  common-ground  circuits  in  the  micro- 
phone lead.  The  ground  connection  for  the 
shielded  cable  which  runs  from  the  transmitter 
to  the  microphone  should  be  made  at  only  one 
point,  preferably  directly  adjacent  to  the  grid 
of  the  first  tube  in  the  speech  amplifier.  The 
use  of  a low-level  microphone  usually  will  re- 
quire the  addition  of  two  speech  stages  (a  pen- 
tode and  a triode),  but  these  stages  will  take 
only  a milliampere  or  two  of  plate  current,  and 
150  ma.  per  tube  of  heater  current. 

PE-103A  Dyna-  Because  of  its  availability 
motor  Power  Unit  on  the  surplus  market  at  a 
low  price  and  its  suitability 
for  use  with  about  as  powerful  a mobile  trans- 
mitter as  can  be  employed  in  a passenger  car 
without  resorting  to  auxiliary  batteries  or  a 
special  generator,  the  PE-103A  is  probably 
the  most  widely  used  dynamotor  for  amateur 
work.  Therefore  some  useful  information  will 
be  given  on  this  unit. 

The  nominal  rating  of  the  unit  is  500  volts 
and  l60  ma.,  but  the  output  voltage  will  of 
course  vary  with  load  and  is  slightly  higher 
with  the  generator  charging.  Actually  the  160 
ma.  rating  is  conservative,  and  about  275  ma. 
can  be  drawn  intermittently  without  overheat- 
ing, and  without  damage  or  excessive  brush 
or  commutator  wear.  At  this  current  the  unit 
should  not  be  run  for  more  than  10  minutes  at 


a time,  and  the  average  **on**  time  should  not 
be  more  than  half  the  average  ”off”  time. 

The  output  voltage  vs.  current  drain  is 
shown  approximately  in  figure  11.  The  exact 
voltage  will  depend  somewhat  upon  the  loss 
resistance  of  the  primary  connecting  cable 
and  whether  or  not  the  battery  is  on  charge. 
The  primary  current  drain  of  the  dynamotor 
proper  (excluding  relays)  is  approximately  16 
amperes  at  100  ma.,  21  amperes  at  160  ma., 
26  amperes  at  200  ma.,  and  31  amperes  at  250 
ma. 

Only  a few  of  the  components  in  the  base 
are  absolutely  necessary  in  an  amateur  mobile 
installation,  and  some  of  them  can  just  as  well 
be  made  an  integral  part  of  the  transmitter  if 
desired.  The  base  can  be  removed  for  salvage 
components  and  hardware,  or  the  dynamotor 
may  be  purchased  without  base. 

To  remove  the  base  proceed  as  follows; 
Loosen  the  four  thumb  screws  on  the  base 
plate  and  remove  the  cover.  Remove  the  four 
screws  holding  the  dynamotor  to  the  base 
plate.  Trace  the  four  wires  coming  out  of  the 
dynamotor  to  their  terminals  and  free  the  lugs. 
Then  these  four  wires  can  be  pulled  through 
the  two  rubber  grommets  in  the  base  plate  when 
the  dynamotor  is  separated  from  the  base  plate. 
It  may  be  necessary  to  bend  the  eyelets  in  the 
large  lugs  in  order  to  force  them  through  the 
gtomm  ets. 

Next  remove  the  two  end  housings  on  the 
dynamotor.  Each  is  held  with  two  screws.  The 
high-voltage  commutator  is  easily  identified 
by  its  narrower  segments  and  larger  diameter. 
Next  to  it  is  the  12-volt  commutator.  The  6- 
volt  commutator  is  at  the  other  end  of  the  ar- 
mature. The  12-volt  brushes  should  be  removed 
when  only  6-volt  operation  is  planned,  in  order 
to  reduce  the  drag. 

If  the  dynamotor  portion  of  the  PE- 10 3 A 
power  unit  is  a Pioneer  type  VS-25  or  a Rus- 
sell type  530-  (most  of  them  are),  the  wires  to 
the  12-volt  brush  holder  terminals  can  be  cross 
connected  to  the  6-volt  brush  holder  terminals 
with  heavy  jumper  wires.  One  of  the  wires  dis- 
connected from  the  12-volt  brush  terminals  is 
the  primary  12-volt  pigtail  and  will  come  free. 
The  other  wire  should  be  connected  to  the  op- 
posite terminal  to  form  one  of  the  jumpers. 

With  this  arrangement  it  is  necessary  only 
to  remove  the  6-volt  brushes  and  replace  the 
12-volt  brushes  in  case  the  6-volt  commutator 
becomes  excessively  dirty  or  worn  or  starts 
throwing  solder.  No  difference  in  output  volt- 
age will  be  noted,  but  as  the  12-volt  brushes 
are  not  as  heavy  as  the  6- volt  brushes  it  is 
not  permissible  ^o  draw  more  than  about  150 
ma.  except  for  emergency  use  until  the  6-volt 
commutator  can  be  turned  down  or  repaired. 


RING 
■ TIP 


SHELL 

(GROUND) 


HANDBOOK 


Noise  Suppression  527 


Figure  11 

APPROXIMATE  OUTPUT  VOLTAGE  VS. 
LOAD  CURRENT  FOR  A PE-103A 
DYNAMOTOR 


At  150  ma.  or  less  the  12-volt  brushes  will 
last  almost  as  long  as  the  6-volt  brushes. 

The  reason  that  these  particular  dynamotors 
can  be  operated  in  this  fashion  is  that  there 
are  two  6-volt  windings  on  the  armature,  and 
for  12-volt  operation  the  two  ate  used  in  series 
with  both  commutators  working.  The  arrange- 
ment described  above  simply  substitutes  for 
the  regular  6-volt  winding  the  winding  and 
commutator  which  ordinarily  came  into  opera- 
tion only  on  12-volt  operation.  Some  operators 
have  reported  that  the  regulation  of  the  PE- 
IO3A  may  be  improved  by  operating  both  com- 
mutators in  parallel  with  the  6-volt  line. 

The  three  wires  now  coming  out  of  the  dy- 
namotor  are  identified  as  follows:  The  smaller 
wire  is  the  positive  high  voltage.  The  heavy 
wire  leaving  the  same  grommet  is  positive  6 
volts  and  negative  high  voltage.  The  single 
heavy  wire  leaving  the  other  grommet  is  nega- 
tive 6 volts.  Whether  the  car  is  positive  or 
negative  ground,  negative  high  voltage  can  be 
taken  as  car-frame  ground.  With  the  negative 
of  the  car  battery  grounded,  the  plate  current 
can  return  through  the  car  battery  and  the  ar- 
mature winding.  This  simply  puts  the  6 volts 
in  series  with  the  500  volts  and  gives  6 extra 
volts  plate  voltage. 

The  trunk  of  a car  gets  very  warm  in  sum- 
mer, and  if  the  transmitter  and  dynamotor  are 
mounted  in  the  trunk  it  is  recommended  that 
the  end  housings  be  left  off  the  dynamotor  to 
facilitate  cooling.  This  is  especially  impor- 
tant in  hot  climates  if  the  dynamotor  is  to  be 
loaded  to  more  than  200  ma. 

When  replacing  brushes  on  a PE-103A  check 
to  see  if  the  brushes  are  marked  negative  and 
positive.  If  so,  be  sure  to  install  them  accord- 
ingly, because  they  ate  not  of  the  same  mater- 
ial. The  dynamotor  will  be  marked  to  show 
which  holder  is  negative. 


When  using  a PE-103A,  or  any  dynamotor 
for  that  matter,  it  may  be  necessary  to  devote 
one  set  of  contacts  on  one  of  tbe  control  re- 
lays to  breaking  the  plate  or  screen  voltage 
to  the  transmitter  oscillator.  If  these  are  sup- 
plied by  the  dynamotot,  because  the  output  of 
a dynamotor  takes  a moment  to  fall  to  zero 
when  the  primary  power  is  removed. 


26-5  Vehicular  Noise 

Suppression 

Satisfactory  reception  on  frequencies  above 
tbe  broadcast  band  usually  requires  greater 
attention  to  noise  suppression  measures.  The 
required  measures  vary  with  the  particular  ve- 
hicle and  the  frequency  range  involved. 

Most  of  the  various  types  of  noise  that  may 
be  present  in  a vehicle  may  be  broken  down 
into  the  following  main  categories; 

( 1)  Ignition  noise. 

(2)  Wheel  static  (tire  static,  brake  static, 
and  intermittent  ground  via  front  wheel  bear- 
ings). 

(3)  "Hash”  from  voltage  regulator  con- 
tacts. 

(4)  "Whine”  from  generator  commutator  seg- 
ment make  and  break. 

(5)  Static  from  scraping  connections  be- 
tween various  parts  of  the  car. 

There  is  no  need  to  suppress  ignition  noise 
completely,  because  at  the  higher  frequencies 
ignition  noise  from  passing  vehicles  makes 
the  use  of  a noise  limiter  mandatory  anyway. 
However,  the  limiter  should  not  be  given  too 
much  work  to  do,  because  at  high  engine 
speeds  a noisy  ignition  system  will  tend  to 
mask  weak  signals,  even  though  with  the  lim- 
iter working,  ignition  "pops”  may  appear  to 
be  completely  eliminated. 

Another  reason  for  good  ignition  suppres- 
sion at  the  source  is  that  strong  ignition 
pulses  contain  enough  energy  when  integrated 
to  block  the  a-v-c  circuit  of  the  receiver,  caus- 
ing the  gain  to  drop  whenever  the  engine  is 
speeded  up.  Since  the  a-v-c  circuits  of  the 
receiver  obtain  no  benefit  from  a noise  clip- 
per, it  is  important  that  ignition  noise  be  sup- 
pressed enough  at  the  source  that  the  a-v-c 
circuits  will  not  be  affected  even  when  the 
engine  is  running  at  high  speed. 

Ignition  Noise  The  following  procedure 
should  be  found  adequate  for 
reducing  tbe  ignition  noise  of  practically  any 
passenger  cat  to  a level  which  the  clipper  can 
handle  satisfactorily  at  any  engine  speed  at 
any  frequency  from  500  kc.  to  148  Me.  Some 


528  Mobile  Equipment 


THE  RADIO 


of  the  measures  may  already  have  been  taken 
when  the  auto  receiver  was  installed. 

First  either  install  a spark  plug  suppressor 
on  each  plug,  or  else  substitute  Autolite  re- 
sistor plugs.  The  latter  are  more  effective  than 
suppressors,  and  on  some  cars  ignition  noise 
is  reduced  to  a satisfactory  level  simply  by 
installing  them.  However,  they  may  not  do  an 
adequate  job  alone  after  they  have  been  in  use 
for  a while,  and  it  is  a good  idea  to  take  the 
following  additional  measures. 

Check  all  high  tension  connections  for  gaps, 
particularly  the  "pinch  fit”  terminal  connec- 
tors widely  used.  Replace  old  high  tension 
wiring  that  may  have  become  leaky. 

Check  to  see  if  any  of  the  high  tension  wir- 
ing is  cabled  with  low  tension  wiring,  or  run 
in  the  same  conduit.  If  so,  reroute  the  low  ten- 
sion wiring  to  provide  as  much  separation  as 
practicable. 

By-pass  to  ground  the  6-volt  wire  from  the 
ignition  coil  to  the  ignition  switch  at  each 
end  with  a O.l-pfd.  molded  case  paper  capaci- 
tor in  parallel  with  a .001-pfd.  mica  or  cer- 
amic, using  the  shortest  possible  leads. 

Check  to  see  that  the  hood  makes  a good 
ground  contact  to  the  car  body  at  several 
points.  Special  grounding  contactors  are  avail- 
able for  attachment  to  the  hood  lacings  on  cars 
that  otherwise  would  present  a gtounding 
problem. 

If  the  high-tension  coil  is  mounted  on  the 
dash,  it  may  be  necessary  to  shield  the  high 
tension  wire  as  far  as  the  bulkhead,  unless 
it  already  is  shielded  with  armored  conduit. 

Wheel  Static  Wheel  static  is  either  static 
electricity  generated  by  rotation 
of  the  tires  and  brake  drums,  or  is  noise  gen- 
erated by  poor  contact  between  the  front 
wheels  and  the  axles  (due  to  the  grease  in  the 
bearings).  The  latter  type  of  noise  seldom  is 
caused  by  the  tear  wheels,  but  tire  static  may 
of  course  be  generated  by  all  four  tires. 

Wheel  static  can  be  eliminated  by  insertion 
of  grounding  springs  under  the  front  hub  caps, 
and  by  inserting  "tire  powder”  in  all  inner 
tubes.  Both  items  are  available  at  radio  parts 
stores  and  from  most  auto  radio  dealers. 

Voltage  Regulator  Certain  voltage  regulators 

Hash  generate  an  objectionable 

amount  of  "hash”  at  the 
higher  frequencies,  particularly  in  the  v-h-f 
range.  A large  by-pass  will  affect  the  operation 
of  the  regulator  and  possibly  damage  the 
points.  A small  by-pass  can  be  used,  however, 
without  causing  trouble.  At  frequencies  above 
the  frequency  at  which  the  hash  becomes  ob- 
jectionable (approximately  20  Me.  or  so)  a 
small  by-pass  is  quite  effective.  A 0.001-pfd. 


mica  capacitor  placed  from  the  field  terminal 
of  the  regulator  to  ground  with  the  shortest 
possible  leads  often  will  produce  sufficient 
improvement.  If  not,  a choke  consisting  of  a- 
bout  60  turns  of  no.  18  d.c.c.  or  bell  wire 
wound  on  a )^-inch  form  can  be  added.  This 
should  be  placed  right  at  the  regulator  termi- 
nal, and  the  0.001-pfd.  by-pass  placed  from 
the  generator  side  of  the  choke  to  ground. 

Generator  Whine  Generator  "whine”  often  can 
be  satisfactorily  suppressed 
from  550  kc.  to  148  Me.  simply  by  by-passing 
the  armature  terminal  to  ground  with  a special 
"auto  radio”  by-pass  of  0.25  or  0.5  pfd.  in 
parallel  with  a 0.001-ftfd.  mica  or  ceramic  ca- 
pacitor. The  former  usually  is  placed  on  the 
generator  when  an  auto  radio  is  installed,  but 
must  be  augmented  by  a mica  or  ceramic  ca- 
pacitor with  short  leads  in  order  to  be  effec- 
tive at  the  higher  frequencies  as  well  as  on  the 
broadcast  band. 

When  more  drastic  measures  are  required, 
special  filters  can  be  obtained  which  are  de- 
signed for  the  purpose.  These  are  recommend- 
ed for  stubborn  cases  when  a wide  frequency 
range  is  involved.  For  reception  only  over  a 
comparatively  narrow  band  of  frequencies, 
such  as  the  10-meter  amateur  band,  a highly 
effective  filter  can  be  improvised  by  connect- 
ing between  the  previously  described  parallel 
by-pass  capacitors  and  the  generator  arma- 
ture termini  a resonant  choke.  This  may  con- 
sist of  no.  10  enamelled  wire  wound  on  a suit- 
able form  and  shunted  with  an  adjustable  trim- 
mer capacitor  to  permit  resonating  the  com- 
bination to  the  center  of  the  frequency  band 
involved.  For  the  lO-meter  band  11  turns  close 
wound  on  a one-inch  form  and  shunted  by  a 
3-30  p/rfd.  compression-type  mica  trimmer  is 
suitable.  The  trimmer  should  be  adjusted  ex- 
perimentally at  the  center  frequency. 

When  generator  whine  shows  up  after  once 
being  satisfactorily  suppressed,  the  condition 
of  the  brushes  and  commutator  should  be 
checked.  Unless  a by-pass  capacitor  has 
opened  up,  excessive  whine  usually  indicates 
that  the  brushes  or  commutator  are  in  need  of 
attention  in  order  to  prevent  damage  to  the 
generator. 

Body  Static  Loose  linkages  or  body  or  frame 
joints  anywhere  in  the  cat  ate 
potential  static  producers  when  the  car  is  in 
motion,  particularly  over  a rough  road.  Locat- 
ing the  source  of  such  noise  is  difficult,  and 
the  simplest  procedure  is  to  give  the  car  a 
thorough  tightening  up  in  the  hope  that  the 
offending  poor  contacts  will  be  caught  by  the 
procedure.  The  use  of  braided  bonding  straps 
between  the  various  sections  of  the  body  of 
the  car  also  may  prove  helpful. 


HANDBOOK 


Noise  Suppression  529 


Miscellaneous  There  are  several  other  poten- 
tial noise  sources  on  a pas- 
senger vehicle,  but  they  do  not  necessarily 
give  trouble  and  therefore  require  attention 
only  in  some  cases. 

The  heat,  oil  pressure,  and  gas  gauges  can 
cause  a rasping  or  scraping  noise.  The  gas 
gauge  is  the  most  likely  offender.  It  will  cause 
trouble  only  when  the  cat  is  rocked  or  is  in 
motion.  The  gauge  units  and  panel  indicators 
should  both  be  by-passed  with  the  0.1-ftfd. 
paper  and  0.001-gfd.  mica  or  ceramic  combina- 
tion previously  described. 

At  high  car  speeds  under  certain  atmospheric 
conditions  corona  static  may  be  encountered 
unless  means  are  taken  to  prevent  it.  The  re- 
ceiving-type auto  whips  which  employ  a plas- 
tic ball  tip  ate  so  provided  in  order  to  minimize 
this  type  of  noise,  which  is  simply  a discharge 
of  the  frictional  static  built  up  on  the  car.  A 
whip  which  ends  in  a relatively  sharp  metal 
point  makes  an  ideal  discharge  point  for  the 
static  charge,  and  will  cause  corona  trouble 
at  a much  lower  voltage  than  if  the  tip  were 
hooded  with  insulation.  A piece  of  Vinylite 
sleeving  slipped  over  the  top  portion  of  the 
whip  and  wrapped  tightly  with  heavy  thread 
will  prevent  this  type  of  static  discharge  un- 
der practically  all  conditions.  An  alternative 
arrangement  is  to  wrap  the  top  portion  of  the 
whip  with  Scotch  brand  electrical  tape. 

Generally  speaking  it  is  undesirable  from 
the  standpoint  of  engine  performance  to  use 
both  spark-plug  suppressors  and  a distributor 
suppressor.  Unless  the  distributor  rotor  clear- 
ance is  excessive,  noise  caused  by  sparking 
of  the  distributor  rotor  will  not  be  so  bad  but 
what  it  can  be  handled  satisfactorily  by  a 
noise  limiter.  If  not,  it  is  preferable  to  shield 
the  hot  lead  between  ignition  coil  and  distri- 
butor rather  than  use  a distributor  suppressor. 


In  many  cases  the  control  rods,  speedometer 
cable,  etc.,  will  pick  up  high-tension  noise 
under  the  hood  and  conduct  it  up  under  the 
dash  where  it  causes  trouble.  If  so,  all  con- 
trol rods  and  cables  should  be  bonded  to  the 
fire  wall  (bulkhead)  where  they  pass  through, 
using  a short  piece  of  heavy  flexible  braid  of 
the  type  used  for  shielding. 

In  some  cases  it  may  be  necessary  to  bond 
the  engine  to  the  frame  at  each  rubber  engine 
mount  in  a similar  maimer.  If  a rear  mounted 
whip  is  employed  the  exhaust  tail  pipe  also 
should  be  bonded  to  the  frame  if  supported  by 
rubber  mounts. 


Locating  Determining  the  source  of  cer- 

Noise  Sources  tain  types  of  noise  is  made 
difficult  when  several  things 
are  contributing  to  the  noise,  because  elimi- 
nation of  one  source  often  will  make  little  or 
no  apparent  difference  in  the  total  noise.  The 
following  procedure  will  help  to  isolate  and 
identify  various  types  of  noise. 

Ignition  noise  will  be  present  only  when  the 
ignition  is  on,  even  though  the  engine  is  turn- 
ing over. 

Generator  noise  will  be  present  when  the 
motor  is  turning  over,  regardless  of  whether 
the  ignition  switch  is  on.  Slipping  the  drive 
belt  off  will  kill  it. 

Gauge  noise  usually  will  be  present  only 
when  the  ignition  switch  is  on  or  in  the  "left” 
position  provided  on  some  cars. 

Wheel  static  when  present  will  persist  when 
the  car  clutch  is  disengaged  and  the  ignition 
switch  turned  off  (or  to  the  left  position),  with 
the  car  coasting. 

Body  noise  will  be  noticeably  worse  on  a 
bumpy  road  than  on  a smooth  toad,  particu- 
larly at  low  speeds. 


CHAPTER  TWENTY-SEVEN 


Receivers  and  Transceivers 


Receiver  construction  has  just  about  become 
a lost  art.  Excellent  general  coverage  receivers 
are  available  on  the  market  in  many  price 
ranges.  However,  even  the  most  modest  of 
these  receivers  is  relatively  expensive,  and  most 
of  the  receivers  are  designed  as  a compromise 
— they  must  suit  the  majority  of  users,  and 
they  must  be  designed  with  an  eye  to  the  price. 

It  is  a tribute  to  the  receiver  manufacturers 
that  they  have  done  as  well  as  they  have.  Even 
so,  the  c-w  man  must  often  pay  for  a high- 
fidelity  audio  system  and  S-meter  he  never 
uses,  and  the  phone  man  must  pay  for  the  c-w 
man’s  crystal  filter.  For  one  amateur,  the  re- 
ceiver has  too  much  bandspread;  for  the  next, 
too  little.  For  economy’s  sake  and  for  ease 


of  alignment,  low-Q  coils  are  often  found  in 
the  r-f  circuits  of  commercial  receivers,  mak- 
ing the  set  a victim  of  cross-talk  and  over- 
loading from  strong  local  signals.  Rarely  does 
the  purchaser  of  a commercial  receiver  realize 
that  he  could  achieve  the  results  he  desires 
in  a home-built  receiver  if  he  left  off  the  frills 
and  trivia  which  he  does  not  need  but  which 
he  must  pay  for  when  he  buys  a commercial 
product. 

The  ardent  experimenter,  however,  needs 
no  such  arguments.  He  builds  his  receiver 
merely  for  the  love  of  the  game,  and  the  thrill 
of  using  a product  of  his  own  creation. 

It  is  hoped  that  the  receiving  equipment  to 
be  described  in  this  chapter  will  awaken  the 


FIGURE  1 

COMPONENT  NOMENCLATURE 


CAPACITORS'. 

1 - VALUES  BELOW  999  JUJUFD  ARE  INDICATED  IN  UNITS. 
EXAMPLE!  iSOJUiJFD  DESIGNATED  AS  150. 

2-  VALUES  ABOVE  999  JUJUFD  ARE  INDICATED  IN  DECIMALS. 
EXAMPLE!  .005JUFD  DESIGNATED  AS  .005. 

3-  OTHER  CAPACITOR  VALUES  ARE  AS  STATED. 

EXAMPLE:  10JUFD,  O.SJUJUFD,  ETC. 

4-  Type  of  capacitor  is  indicated  beneath  the  value 

DESIGNATION. 

SM=  SILVER  MICA 
C = CERAMIC 
M=  MICA 
P = PAPER 


EXAMPLE! 


250  .01 


.001 


VOLTAGE  RATING  OF  ELECTROLYTIC  OR  "FILTER* 
CAPACITOR’  IS  INDICATED  BELOW  CAPACITY  DESIGNATION. 


6-  THE  CURVED  LINE  IN  CAPACITOR  SYMBOL  REPRESENTS 
THE  OUTSIDE  FOIL  "GROUND"  OF  PAPER  CAPACITORS, 

THE  NEGATIVE  ELECTRODE  OF  ELECTROLYTIC  CAPACITORS, 
OR  THE  ROTOR  OF  VARIABLE  CAPACITORS. 


RESISTORS- 

1-  RESISTANCE  VALUES  ARE  STATED  IN  OHMS,  THOUSANDS 
OF  OHMS  (K).  AND  MEGOHMS  (M  ). 

EXAMPLE!  Z70  OHMS  = 270 
4700  OHMS  = 4.7  K 
33,000  OHMS  = 33  K 
100,000  OHMS  - 100  K OR  0.1  M 
33,000,000  0HMS=  33  M 

2-  ALL  RESISTORS  ARE  1- WATT  COMPOS ITION  TYPE  UNLESS 
OTHERWISE  NOTED.  WATTAGE  NOTATION  IS  THEN  INDICATED 
BELOW  RESISTANCE  VALUE. 

47  K 


EXAMPLE: 


O.S 


INDUCTORS: 

MlCROHENRIES  = Ji;H 
MILLI  HENRIES  = MH 
HENRIES  = H 


SCHEMATIC  SYMBOLS: 

OP  1 t 


CONDUCTORS  JOINED 


CONDUCTORS  CROSSING 
BUT  NOT  JOINED 


X 

CHASSIS  GROUND 


530 


531 


experimenter’s  instinct,  even  in  those  individ- 
uals owning  expensive  commercial  receivers- 
These  lucky  persons  have  the  advantage  of 
comparing  their  home-built  product  against  the 
best  the  commercial  market  has  to  offer.  Some- 


times such  a comparison  is  surprising. 

When  the  builder  has  finished  the  wiring  of 
a receiver  it  is  suggested  that  he  check  his 
wiring  and  connections  carefully  for  possible 
errors  before  any  voltages  are  applied  to  the 


TUBULAR  CAPACITOR 
NORMALLY  STAMPED 
FOR  VALUE 

■aND  I SIGNIFICANT  FIGURE 
MULTIPLIER 


SIG.  VOLTAGE  FIG. 


A Z- DIGIT  VOLTAGE  RATING  INDICATES  MORE  THAN 
900  V.  ADD  Z ZEROS  TO  END  OF  Z DIGIT  NUMBER. 


TOLERANCE 


MOLDED  FLAT  CAPACITOR 

COMMERCIAL  CODE 


[-WORKING  VOLTS 


MULTIPLIER 


IGNIFtCANT 

GURE 


JAN  CODE  CAPACITOR 
SILVER  p— 1^ST|  significant  fig, 

^MULTIPLIER 

-TOLERANCE 
-CHARACTERISTIC 


Figure  2 

STANDARD  COLOR  CODE  FOR  RESISTORS  AND  CAPACITORS 

The  standard  color  code  provides  the  necessary  information  required  to  properly  identify  color  coded 
resistors  and  capacitors.  Refer  to  the  color  code  for  numerical  values  and  the  number  of  zeros  (or  multi^ 
plier)  assigned  to  the  colors  used.  A fourth  color  band  on  resistors  determines  the  tolerance  rating  as 
follows:  Gold =5%,  silver  = 10%.  Absence  of  the  fourth  band  indicates  a 20%  tolerance  rating. 
Tolerance  rating  of  capacitors  is  determined  by  the  color  code.  For  example:  Red  — 2%,  green  = 5%,  etc. 
The  voltage  rating  of  capacitors  is  obtained  by  multiplying  the  color  value  by  100.  For  example: 

Orange=3X  100,  or  300  volts. 


532  Receivers  and  Transceivers 


THE  RADIO 


FIGURE  3 

COMPONENT  COLOR  CODING 

POWER  TRANSFORMERS 

PRIMARY  LEADS 

— BLACK 

/F  TAPPED: 

TAP 

— BLACKyyELLOW 

end 

— BLACK  y RED 

HIGH  VOLTAGE  WINDING  

— RED 

CENTER- TAP 

— REOyyELLOW 

RECTIFIER  FILAMENT  WINDING 

— YELLOW 

CENTER- TAP 

— YELLOWyBLUE 

FILAMENT  WINDING  N*  1 

— BREEN 

CENTER- TAP 

— BREENyYEL  L OW 

FILAMENT  WINDING  N»  2 

BROWN 

CENTER- TAP 

— BROWNy  YELLOW 

FILAMENT  WINDING  N«  3 

— SLATE 

CENTER-TAP 

SLATE  yYELLOW 

I-F  TRANSFORMERS 

PLATE  LEAD 

— BLUE 

a+  LEAD 

— RED 

GRID  (OPOtODE)  LEAD 

— BREEN 

A-V-C  {OR  CPOUNO)  LEAD 

— BLACK 

AUDIO  TRANSFORMERS 

PLATE  LEAD  {PR/.) 

— BLUE  OR  BROWN 

0+-  LEAD  {PR/.  ) 

— RED 

GRID  LEAD  {SEC.) 

— BREEN  OR  YELLOW 

GRID  RETURN  {SEC.  ) 

— BLACK 

circuits.  If  possible,  the  wiring  should  be 
checked  by  a second  party  as  a safety  mea- 
sure. Some  tubes  can  be  permanently  damaged 
by  having  the  wrong  voltages  applied  to  their 
electrodes.  Electrolytic  capacitors  can  be 
ruined  by  hooking  them  up  with  the  wrong 
voltage  polarity  across  the  capacitor  terminals. 
Transformer,  choke  and  coil  windings  may  be 


Figure  4 

TWO  TRANSISTOR  BROADCAST 
RECEIVER  IS  EXCELLENT 
BEGINNER'S  PROJECT 

This  simple  receiver  works  from  a single  9- 
volt  battery,  providing  many  hours  of 
pleasant  listening.  Two  transistors  and  a 
diode  in  a reflex  circuit  provide  ample 
loudspeaker  volume.  Tuning  control  is  at 
left,  with  audio  control  beneath  it.  Midget 
speaker  is  at  right.  Receiver  is  housed  in 
plastic  case. 


damaged  by  incorrect  wiring  of  the  high-volt- 
age leads. 

The  problem  of  meeting  and  overcoming 
such  obstacles  is  just  part  of  the  game.  A true 
radio  amateur  ( as  opposed  to  an  amateur 
broadcaster)  should  have  adequate  knowledge 
of  the  art  of  communication.  He  should  know 
quite  a bit  about  his  equipment  (even  if  pur- 
chased) and,  if  circumstances  permit,  he  should 
build  a portion  of  his  own  equipment.  Those 
amateurs  that  do  such  construction  work  are 
convinced  that  half  of  the  enjoyment  of  the 
hobby  may  be  obtained  from  the  satisfaction 


Figure  5 


SCHEMATIC  OF  TRANSISTOR  BROADCAST  RECEIVER 

B — 9-volt  transistor  battery.  RCA  VS-300 

Cl— 365  /x/ifd.  Lafayette  Radio  Co.  MS-214,  or  Allied  Radio  Co.  61H-009 
Li — Transistor  *‘Loopstick"  coil.  Lafayette  Radio  Co.  MS-166 
LS — 3"  loudspeaker,  12-ohm  voice  coil.  Lafayette  Radio  Co.  SIC-39 
Ti — 500  ohm  pri.,  12  ohm  sec.  Transistor  transformer.  Thordarson  TR-18 


HANDBOOK 


Circuitry  533 


of  building  and  operating  their  own  receiving 
and  transmitting  equipment. 

The  Transceiver  A popular  item  of  equip- 
ment on  "five  meters” 
during  the  late  "thirties,”  the  transceiver  is 
making  a comeback  today  complete  with  mod- 
ern tubes  and  circuitry.  In  brief,  the  transceiver 
is  a packaged  radio  station  combining  the  ele- 
ments of  the  receiver  and  transmitter  into  a 
single  unit  having  a common  power  supply 
and  audio  system.  The  present  trend  roward 
compact  equipment  and  the  continued  growth 
of  single  sideband  techniques  combine  natural- 
ly with  the  space-saving  economies  of  the 
transceiver.  Various  transceiver  circuirs  for  the 
higher  frequency  amateur  bands  are  shown  in 
this  chapter.  The  experimenter  can  start  from 
these  simple  circuits  and  using  modern  minia- 
ture tubes  and  components  can  design  and 
build  his  complete  station  in  a cabinet  no  larg- 
er than  a pre-war  receiver. 

27-1  Circuitry  and 
Components 

It  is  the  practice  of  the  editors  of  this 
Handbook  to  place  as  much  usable  information 
in  the  schematic  illustration  as  possible.  In 
order  to  simplify  the  drawing  the  component 
nomenclature  of  figure  1 is  used  in  all  the  fol- 
lowing construction  chapters. 

The  electrical  value  of  many  small  circuit 
components  such  as  resistors  and  capacitors  is 
often  indicated  by  a series  of  colored  bands 


or  spots  placed  on  the  body  of  the  component. 
Several  color  codes  have  been  used  in  the  past, 
and  are  being  used  in  modified  form  in  the 
present  to  indicate  component  values.  The 
most  important  of  these  color  codes  are  illus- 
trated in  figure  2.  Other  radio  components 
such  as  power  transformers,  i-f  transformers, 
chokes,  etc.  have  their  leads  color-coded  for 
easy  identification  as  tabulated  in  figure  3. 

27-2  A Simple 

Transistorized  Portable 
B-C  Receiver 

Illustrated  in  figures  4,  5,  and  6 is  an  easy 
to  construct  two  transistor  portable  broadcast 
receiver  that  is  an  excellent  circuit  for  the 
beginner  to  build.  The  receiver  covers  the 
range  of  500  kc.  to  1500  kc.  and  needs  no 
external  antenna  when  used  close  to  a high 
power  broadcasting  station.  An  external  anten- 
na may  be  added  for  more  distant  reception. 
The  receiver  is  powered  from  a single  9-volt 
miniature  transistor  battery  and  delivers  good 
speaker  volume,  yet  draws  a minimum  of  cur- 
rent permitting  good  battery  life. 

Circuit  Operation  of  the  receiver  may 

Description  be  understood  by  referring  to 
the  schematic  diagram  of  figure 
5.  The  tuned  circuit  Li-G  resonates  at  the 
frequency  of  the  broadcasting  station.  A por- 
tion of  the  r-f  energy  is  applied  to  the  base  of 
the  2N112  p-n-p  type  transistor.  A tapped 


Figure  6 

INTERIOR  VIEW 
OF  TRANSISTOR 
RECEIVER 

The  speaker  and  output 
transformer  are  mounted 
at  the  left  of  the  Ma- 
sonite chassis.  Top,  cen- 
ter is  the  ^'loopsiick"  r-/ 
coil,  and  directly  to  the 
right  is  the  10  millihenry 
r-f  choke  in  the  collector 
lead  of  the  2N112  trans- 
istor. Battery  is  at  tower 
right. 


534  Receivers  and  Transceivers 


THE  RADIO 


winding  is  placed  on  coil  Li  to  achieve  an  im- 
pedance match  to  the  low  base  impedance  of 
the  transistor.  Emitter  bias  is  used  on  this 
stage,  and  the  amplified  signal  is  capacity- 
coupled  from  the  collector  circuit  to  a 1N34 
diode  rectifier.  The  rectifier  audio  signal  is 
recovered  across  the  2K  diode  load  resistor, 
which  takes  the  form  of  the  audio  volume 
control  of  the  receiver.  The  diode  operates  in 
an  untuned  circuit,  the  selectivity  of  the  re- 
ceiver being  determined  by  the  tuned  circuit 
in  the  r-f  amplifier  stage. 

The  audio  signal  taken  from  the  arm  of  the 
volume  control  (Ri)  is  applied  to  the  base  of 
the  2N112  r-f  amplifier  which  functions  sim- 
ultaneously as  an  audio  amplifier  stage-  The 
amplified  audio  signal  is  recovered  across  the 
2K  collector  load  resistor  of  the  2N112,  and 
is  capacitively  coupled  to  the  base  of  a 2N217 
p-n-p  audio  transistor.  This  stage  is  base  and 
emitter  biased,  having  the  output  transformer 
in  the  collector  circuit.  Optimum  collector  load 
for  the  2N217  is  approximately  500  ohms, 
and  the  2N217  develops  a maximum  audio 
signal  of  75  milliwatts  at  this  load  impedance. 
Transformer  Ti  matches  the  transistor  circuit 
to  the  12  ohm  miniature  loudspeaker.  The 
receiver  draws  a maximum  signal  current  of 
11  milliamperes  from  the  9-volt  battery  sup- 
ply. 


Receiver  The  complete  receiver  is 

Construction  mounted  upon  a thin  Maso- 

nite board  measuring  IVs"  tt. 
4yg''  in  size.  The  board  fits  within  a small 
plastic  box  having  a hinged  lid  and  handle. 
The  dimensions  of  the  box  are  71/2"  x 5"  x 
lYs",  exclusive  of  the  handle.  Placement  of  the 
major  components  upon  the  board  may  be  seen 
in  figure  6.  The  "loopstick”  antenna  coil  is 
mounted  to  the  top  center  portion  of  the  board 
so  that  it  fits  directly  under  the  handle  of  the 
box.  Slightly  below  and  to  the  right  of  the 
coil  is  the  tuning  capacitor,  Ci.  The  midget 
speaker  is  mounted  in  a cut-out  in  the  board 
at  the  lower  left  corner  with  the  output  trans- 
former Ti  directly  above  it.  The  volume  con- 
trol Ri  is  placed  directly  below  the  tuning 
capacitor.  The  battery  clip  for  the  9-volt  cell 
is  placed  at  the  lower  right  corner  of  the  board. 
A three  terminal  phenolic  strip  is  attached  to 
the  board  above  the  battery  clip.  The  leads  of 
the  2N112  transistor  are  soldered  to  this  strip. 

Before  the  components  are  mounted  on  the 
board,  it  is  placed  within  the  plastic  box  and 
used  as  a drilling  template  for  the  box.  A 
drill-press  and  "fly-cutter”  may  be  employed 
to  cut  the  speaker  hole  in  both  the  board  and 
the  box.  The  plastic  box  should  be  drilled 
slowly  to  prevent  excessive  heat  generation 
and  consequent  melting  of  the  material. 


Figure  7 

BANDWIDTH  CURVES  OF  VARIOUS  l-F  SYSTEMS 

Bandwidth  curves  showing:  A — Seiectiviiy  range  of  most  medium-priced  single-conversion  receivers  with 
crystal  filter  out  of  circuit;  B — Ideal  selectivity  curve  for  voice  reception;  and  C — Selectivity  curve  of  a 
455  kc.  mechanical  filter  with  a 3,1  kilocycle  bandwidth. 


HANDBOOK 


Mechanical  Filter  535 


The  2N217  transistor  is  mounted  upon  a 
second  three  terminal  phenolic  strip  placed  be- 
tween the  tuning  capacitor  and  the  loud  speak- 
er. Use  of  a composition-type  chassis  precludes 
the  employment  of  the  chassis  as  a common 
ground.  A simple  bare  copper  wire  is  therefore 
used  as  a common  "ground”  return.  This  lead 
is  actually  the  "B-plus  Bus,”  shown  in  figure  5. 
The  lead  runs  from  the  positive  terminal  of 
the  battery  clip,  along  the  lower  edge  of  the 
chassis  then  up  to  the  "ground”  terminal  of 
the  "loopstick”  coil,  Li.  All  components  that 
normally  ground  to  the  metal  chassis  in  a 
conventional  assembly  are  attached  to  this  bus. 
Be  sure  to  observe  the  polarity  of  the  electro- 
lytic capacitors  and  the  dry  cell.  Improper  con- 
nection of  these  components  can  damage  the 
transistors  or  other  parts  of  the  circuit. 

The  transistors  are  the  last  items  to  be  con- 
nected into  the  assembly.  Since  the  transistors 
are  heat  sensitive,  it  is  necessary  to  protect  them 
during  the  soldering  operation.  Grasp  the  tran- 
sistor lead  to  be  soldered  with  a long-nose 
pliers  between  the  body  of  the  transistor  and 
the  end  of  the  lead.  The  thermal  capacity  of 
the  metal  pliers  will  act  as  a heat  sink,  pre- 
venting the  heat  of  the  iron  from  damaging 
the  transistor.  Be  sure  to  hold  the  pliers  in 
place  until  the  completed  joint  has  had  time 
to  cool.  When  all  wiring  is  completed  and 
checked,  the  9-volt  battery  can  be  inserted  in 
the  clip  (observe  polarity!)  and  the  set  can 
be  turned  on  by  means  of  switch  Si.  A 0-100 
milliampere  d-c  meter  placed  temporarily 
across  the  open  contacts  of  switch  Si  can  be 
used  to  check  the  transistor  operating  current. 
The  receiver  should  play  immediately  upon 
closing  Si  as  the  transistors  have  no  warm-up 
time.  If  no  external  antenna  is  used,  the  re- 
ceiver should  be  moved  about  to  orient  the 
"loopstick”  coil  Li  for  best  pickup  of  each 
individual  broadcast  station.  Adjacent  channel 
interference  can  often  be  eliminated  by  care- 
ful rotation  of  the  set  to  "null  out”  the  offend- 
ing signal.  Ample  loudspeaker  volume  will  be 
obtained  from  local  stations  without  the  use 
of  an  external  antenna. 


Figure  8 

455  KC.  MECHANICAL 
FILTER  ADAPTER 

Enjoy  optimum  selectivity  and  performance 
from  your  present  receiver  by  plugging  in 
this  simple  adapter  that  replaces  the  first 
i-f  amplifier  tube.  The  two  6BJ6  tubes  are 
at  opposite  ends  of  the  chassis  with  the 
mechanical  filter  between  them.  Adapter 
plug  to  fit  i~f  tube  socket  is  shown  in 
foreground. 


27-3  A 455  Kc. 

Mechanical 
Filter  Adapter 

Enjoy  modern-style  selectivity  and  perform- 
ance with  your  present  receiver  by  plugging  in 
this  simple  mechanical  filter  .adapter  that  re- 
places the  first  i-f  tube! 

Many  modern  medium-priced  and  older 
high  priced  communications  receivers  now  in 
general  use  are  convenient  to  operate.  The  re- 
ceivers have  good  frequency  stability  and  ade- 
quate sensitivity,  but  lack  sufficient  "skirt” 
selectivity  to  reject  strong  signals  that  are  only 
a few  kilocycles  higher  or  lower  in  frequency 
from  a desired  signal.  The  shaded  area  of  curve 
A,  figure  7 shows  the  typical  selectivity  char- 
acteristic of  popular  medium-priced  commun- 
ications receivers.  Although  the  "nose”  of  this 
curve  is  usually  only  a few  kilocycles  wide,  the 
"skirt”  selectivity  60  decibels  down  from  the 
peak  may  be  from  15  to  30  kilocycles  broad! 
Small  wonder  that  strong  local  signals  a few 
kilocycles  up  the  band  from  the  station  you 
are  trying  to  copy  may  sometimes  paralyze 
your  receiver. 

It  is  possible  to  add  a new  order  of  selec- 
tivity to  your  present  set  by  constructing  this 
mechanical  filter  adapter  unit,  that  is  substi- 


536  Receivers  and  Transceivers 


THE  RADIO 


Figure  9 

CONNECTIONS  BETWEEN 
MECHANICAL  FILTER, 

ADAPTER  AND  RECEIVER 

Adapter  makes  connections  to  first  i-f 
socket  of  receiver.  Power  may  atso  be  ob- 
tained from  the  receiver. 

mted  for  the  first  455  kc.  intermediate  fre- 
quency amplifier  tube,  without  making  any 
under-chassis  changes  in  the  receiver.  Simply 
connect  the  adapter  to  a power  source  (which 
may  be  your  receiver),  remove  the  first  i-f 
amplifier  tube,  and  insert  two  short  coaxial 
cables  into  the  empty  tube  socket,  as  shown 
in  figure  9.  These  cables  carry  the  i-f  signal 
to  and  from  the  adapter,  which  ate  then  tucked 
away  in  an  unoccupied  corner  of  your  receiver 
or  speaker  cabinet. 

Selectivity  There  are  two  systems  gen- 

Considerations  erally  used  to  obtain  a 
bandpass  characteristic  that 
approaches  the  ideal  communications  selectivity 
curve  for  voice  modulated  signals  shown  at  B 
in  figure  7.  One  system  utilizes  a package  fil- 


ter such  as  the  mechanical  filter  or  a crystal 
lattice  filter.  The  second  system  utilizes  a string 
of  high  Q tuned  circuits  in  the  i-f  amplifier 
to  achieve  the  desired  passband.  This  latter  sys- 
tem can  be  space  consuming,  difficult  to  ad- 
just, and  expensive  if  quality  components  are 
employed. 

On  the  other  hand,  the  mechanical  filter  is 
very  compact.  It  is  readily  available  in  a var- 
iety of  bandwidths,  has  an  excellent  selectivity 
curve,  and  is  roughly  equivalent  in  cost  to 
other  systems  having  comparable  selectivity. 
Curve  C of  figure  7 illustrates  the  selectivity 
curve  of  the  3.1  kilocycle  Collins  mechanical 
filter  which  is  suitable  for  A-M  phone  and 
single  sideband  reception. 

Circuit-  The  complete  circuit  of  the 

Description  mechanical  filter  adapter  is 
shown  in  figure  10.  Two  6BJ6 
pentode  tubes  are  employed  in  the  ciircuit  to 
compensate  for  the  insertion  loss  of  the  filter. 
The  adapter  picks  up  the  455  kc.  signal  from 
the  control  grid  pin  of  the  receiver’s  first  i-f 
amplifier  tube  socket  through  coupling  capaci- 
tor, Cs.  The  signal  passes  through  a coaxial 
line  to  the  grid  of  the  6BJ6  amplifier  tube, 
Vi.  The  plate  circuit  of  Vi  is  capacity  coupled 
to  the  input  terminals  of  the  mechanical  filter 
to  keep  plate  current  from  flowing  through 
the  filter  coil.  A much  wider  signal  voltage 
range  can  be  handled  by  the  filter  without 


Figure  10 

SCHEMATIC  OF  MECHANICAL  FILTER  ADAPTER 
Ct,  Cl — 600  piifd.  ceramic  (270-  and  330  pfifd.  in  parallel) 

Cl,  Cl — 720  /ififd.  ceramic 

Cs,  Cl — 70  iLiifd.  tubular  ceramic  (Aeravox  CI-1) 

FL — 455  kc.  mechanical  filter  (3.1  kc.  bandwidth)  (Collins  4SSJ-3I) 

Li,  Li — 200  microhenry  siug-tuned  "loop  stick"  coil  (Superex  VL,  or  J.  W.  Miller  6300) 

P] — Mole  octal  plug  with  retaining  ring  (Amphenol  86-PM8) 

Ri — 18K,  10  watt.  Place  between  pins  4 and  8 of  PI  socket  for  plate  supply  greater  than  130  volts. 
Use  jumper  for  130  volts  or  less  (see  text).  Ri  is  screen  voltage  dropping  rtsistor. 


HANDBOOK 


Mechanical  Filter  537 


V2,  6BJ6 


Figure  1 1 

OPTIONAL  A-V-C  CONNECTIONS 
FOR  THE  ADAPTER 


overloading  when  no  plate  current  flows 
through  the  input  coil.  Both  filter  coils  are 
tuned  to  resonance  at  the  intermediate  fre- 
quency by  fixed  capacitors  G and  Cs. 

The  output  terminals  of  the  mechanical  fil- 
ter are  connected  directly  to  the  input  circuit 
of  the  second  6BJ6  tube.  The  output  signal 
from  this  tube  is  again  capacitively  coupled 
back  into  the  plate  circuit  of  the  first  i-f  stage 
of  the  receiver  through  a second  coaxial  line 
and  coupling  capacitor  G.  The  455  kc.  tuned 
circuits  connected  to  the  plate  of  Vi  and  V2 
are  composed  of  ''loopstick”-type  coils  Li  and 
L:,  shunted  by  fixed  capacitors  Ci  and  C4. 

The  input  and  output  coaxial  cables  are  16- 
inch  lengths  of  RG-58/U  line.  This  cable 
forms  a 40  fj-iJ-d.  ground  leg  of  a capacitor 
voltage  divider  (G  being  the  other  leg)  that 
reduces  the  signal  voltage  applied  to  Vi  to 
about  one-quarter  of  the  voltage  developed 
across  the  secondary  of  the  first  i-f  transformer 
of  the  receiver. 

The  overall  signal  amplification  of  the  adap- 
ter has  been  held  down  to  a few  decibels 
more  than  the  10  db  insertion  loss  of  the  filter 


through  the  use  of  small  input  and  output 
coupling  capacitors  and  large  cathode  bias  re- 
sistors in  both  amplifier  stages.  This  is  suit- 
able for  receivers  having  two  or  more  i-f  am- 
plifier stages,  but  additional  gain  from  the 
adapter  may  be  obtained  by  reducing  the  value 
of  one  or  both  cathode  resistors  to  270  ohms. 
This  may  be  desirable  when  the  adapter  is  op- 
erated with  a receiver  having  only  one  i-f 
amplifier  stage. 

Power  is  brought  into  the  adapter  unit 
through  an  eight-pin  male  plug.  Connections 
to  this  plug  are  made  so  that  the  plug  may  be 
inserted  into  the  NBFM  adapter  socket  on  cer- 
tain National  receiver  models.  Many  commun- 
ications receivers  have  an  accessory  power  sock- 
et on  the  rear  of  the  chassis  from  which 
power  may  be  obtained.  Adapter  requirements 
are  6.3  volts  at  0.3  amperes,  and  105-250 
volts  at  10  milliamperes.  This  is  little  more 
than  the  power  requirements  of  the  i-f  tube 
replaced  by  the  adapter.  A 250-volt  supply 
source  requites  that  an  18K,  12-watt  screen 
voltage  dropping  resistor  (Ri)  be  connected 
between  pins  4 and  8 of  the  power  plug  re- 
ceptacle. When  the  d-c  supply  is  130-volts  or 
less,  a jumper  may  be  placed  directly  between 
pins  4 and  8 (see  figure  10). 

A method  of  applying  automatic  volume 
control  voltage  from  the  receiver  to  the  second 
amplifier  stage  in  the  adapter  is  shown  in 
figure  11.  This  modification  is  mainly  useful 
when  the  adapter  is  connected  to  a receiver 
that  has  few  a-v-c  controlled  stages.  The  auto- 
matic volume  control  voltage  is  taken  from  the 
control  grid  pin  of  the  i-f  amplifier  tube  socket 
through  a simple  R-C  filter  and  is  applied  to 
the  control  grid  of  V2  through  the  output  coil 
of  the  mechanical  filter. 


2^*X  4"  MINIBOX  (.BUD  CU~300^  OR  EQUIVALENT  ) 


Figure  12 

CHASSIS  DRILLING  TEMPLATE  FOR  ADAPTER 

A;  #32  drill,  B:  9/32"  drill,  C:  5/ 8"  diameter  socket  punch,  D:  3/4"  diameter  socket  punch 


538  Receivers  and  Transceivers 


THE  RADIO 


ConstrucHon  The  adapter  is  constructed 

of  Ae  Adapter  on  an  aluminum  box  mea- 
suring 214"  X 2 14  " X 4",  a 
good  compromise  size  that  is  compact,  yet  not 
too  small  for  easy  wiring.  The  primary  design 
and  construction  consideration  of  this  adapter 
is  to  completely  isolate  the  input  and  output 
circuits.  Any  stray  coupling  can  cause  signal 
leakage  around  the  filter  unit,  thus  imparing 
its  effectiveness.  The  construction  shown  al- 
lows maximum  isolation,  yet  permits  sim- 
plicity of  construction.  A drilling  template  for 
the  box  is  given  in  figure  12,  and  the  place- 
ment of  the  major  components  may  be  seen  in 
figures  8,  13,  and  14.  Note  that  the  input 
and  output  circuits  are  isolated  within  the  box 
by  a shield  passing  across  the  center  of  the 
filter  socket. 

After  drilling  and  punching  all  holes,  the 
tube  and  mechanical  filter  sockets,  power  plug, 
and  rubber  grommets  may  be  assembled.  Place 
solder  lugs  under  all  socket  retaining  bolts  for 
ground  connections.  Before  any  wiring  is  done 
a 3"  X 3"  piece  of  perforated  aluminum 
{Reynold’s  "Do-it-yourself"  aluminum,  avail- 
able at  large  hardware  stores)  is  formed  into 
the  shield  shown  in  figures  13  and  14.  A §4” 
flange  is  bent  along  all  edges  of  this  shield 
except  where  it  crosses  the  center  of  the  9-pin 
filter  socket.  A small  notch  is  cut  in  the  shield 
next  to  the  socket  to  pass  the  heater  and  plate 
power  leads  to  the  Va  socket.  The  shield  passes 
across  the  filter  socket  between  pins  3 and  4, 
and  8 and  9,  then  is  bolted  to  a soldering  lug 
that  has  been  soldered  to  pin  1 of  the  filter 
socket.  The  flange  is  also  bolted  to  the  top 
of  the  box  directly  above  coil  La,  and  two  self- 
tapping sheet  metal  screws  are  driven  into  the 
side  flanges  of  the  shield  when  the  other  half 
of  the  box  is  assembled. 


Cable  Constructing  the  two  i-f  tube 

Assembly  socket  cables  takes  little  time  if 
the  assembly  shown  in  figure  15 
is  followed.  First,  cut  two  lengths  of  RG-58/U 
coaxial  cable  1 7 inches  long  and  remove  1 Vi 
inches  of  the  vinyl  cover  on  one  end  of  each 
piece.  Slide  the  braided  outer  shield  back  over 
the  outer  cover  and  trim  the  insulation  and 
center  conductor  so  that  V2  inch  protrudes  be- 
yond the  shield.  Next  cut  the  insulation  to 
expose  54  inch  of  the  center  conductor  and 
trim  one  lead  of  the  coupling  capacitor  Cs  sol- 
dering it  to  the  center  conductor  with  a 54  inch 
overlap.  Cut  narrow  strips  of  plastic  insulating 
tape  and  wrap  them  around  this  joint  up  to 
the  body  diameter  of  the  capacitor  as  shown 
in  figure  15.  Finally,  slide  the  braided  shield 
over  the  capacitor,  pull  it  tight  and  wrap  a 
short  length  of  tinned  copper  wire  twice 
around  the  middle  of  the  capacitor.  Solder  the 
wire  to  the  shield  and  trim  off  the  excess 
shielding.  The  tinned  wires  from  each  cable 
are  then  soldered  to  a pin  taken  from  an  octal 
tube  base,  making  a plug-in  ground  connec- 
tion as  shown  in  figure  8.  Similar  tube  base 
pins  are  soldered  to  the  other  leads,  and  the 
excess  lead  trimmed  off.  The  exposed  cable 
shield  is  then  wrapped  with  plastic  tape. 

If  the  receiver  has  a 7-pin  miniature  tube 
in  the  first  i-f  amplifier  stage,  short  lengths  of 
#18  tinned  wire  may  be  used  for  the  plug-in 
pins  on  the  cables,  or  the  capacitors  and 
ground  lead  may  be  soldered  to  a special  7-pln 
miniature  male  adapter  plug  (Vector  P-7). 

Adopter  For  easy  parts  assembly,  the  filter 
Wiring  shield  may  be  temporarily  re- 
moved, and  replaced  when  wiring 
is  completed.  Heater,  screen  and  plate  power 
wires  are  installed  first,  keeping  all  such  leads 
close  to  the  box  wherever  possible  to  minimize 


Figure  1 3 
UNDERCHASSIS 
VIEW  OF  THE 
ADAPTER 

Slug-iuned  coil  Li  h at 
left,  below  the  second 
stage  6 B J 6 socket. 
Screen  bypass  capacitor 
is  placed  across  socket. 
Capacitors  Cz  and  Cj  may 
be  seen  on  each  side  of 
the  center  shield.  Coil  Li 
is  at  right  of  shield.  In- 
put  cable  is  at  far  right. 
Output  cable  passes 
through  rear  wall  of 
chassis  at  left.  Power 
leads  pass  through  center 
of  shield. 


HANDBOOK 


Mechanical  Filter  539 


stray  signal  pickup.  The  smaller  components 
are  now  soldered  in  place,  after  which  the  co- 
axial cable  input  and  output  leads  are  con- 
nected. About  Ys  inch  of  the  outer  vinyl  jacket 
is  removed  from  the  cables  and  the  shield 
braid  is  twisted  into  a single  conductor.  The 
cable  ends  are  brought  into  the  box  through 
rubber  grommets.  The  twisted  cable  shield  is 
soldered  to  a nearby  ground  lug  and  the  cen- 
ter conductor  is  soldered  to  the  correct  tube 
socket  pin.  Finally  coils  Li  and  La  and  their 
associated  capacitors  are  placed  in  position  and 
wired  into  the  circuit. 

Alignment  The  adapter  is  connected  to 

of  the  Adopter  to  the  communications  re- 
ceiver as  previously  describ- 
ed, following  a wiring  and  power  check  to  en- 
sure that  the  correct  voltages  are  applied  to  the 
various  tube  elements.  The  receiver  should 
then  be  tuned  to  the  center  of  a strong,  steady 
local  signal.  If  the  receiver  has  an  S-meter, 
the  a-v-c  should  be  left  on.  The  slugs  of  coils 
Li  and  Li  are  now  tuned  for  maximum  carrier 
strength  reading  on  the  S-meter.  If  the  receiver 
has  no  S-meter,  a modulated  signal  from  a 
test  oscillator  must  be  used  in  conjunction  with 
a vacuum  tube  voltmeter  placed  across  the 
audio  output  circuit  of  the  receiver.  Tuning 
adjustments  on  the  first  and  second  i-f  trans- 
formers in  the  receiver  are  also  recommended 
for  maximum  output  of  the  adapter. 


POLYETHYLENE 
INNER  INSULATION 


Figure  15 

CROSS-SECTION  VIEW 
OF  SIGNAL  CABLES 


Receiver  Tuning  A somewhat  different  tun- 

Techniques  ing  technique  should  be 

used  for  tuning  A-M  and 
single  sideband  signals  on  a receiver  following 
installation  of  the  mechanical  filter  adapter. 
Modulated  signals  with  carrier  should  be  tuned 
in  so  that  the  carrier  is  placed  on  one  edge, 
rather  than  in  the  center  of  the  i-f  passband 
( figure  7 ) . Only  one  sideband  of  an  A-M 
signal  will  be  heard  at  one  time,  and  the  re- 
ceiver tuning  may  be  shifted  so  that  the  side- 
band on  which  a heterodyne  is  present  may  be 
"pushed  off  the  edge’’  of  the  bandpass  curve 
of  the  i-f  system. 

■When  receiving  single  sideband  or  c-w  sig- 
nals the  beat  frequency  oscillator  of  the  re- 
ceiver is  turned  on  and  adjusted  so  that  the 
frequency  of  the  b-f-o  is  near  one  edge  of  the 


Figure  14 

OBLIQUE  VIEW  OF 
ADAPTER  CHASSIS 

Coil  Li  is  at  left  near 
power  plug,  with  ceil  Li 
mounted  horizontally  at 
right.  Note  Ci  and  C, 
ore  placed  directly  across 
pins  of  mechanical  crys- 
tal  socket. 


Figure  16 

FRONT  VIEW  OF  AMATEUR 
BAND  RECEIVER 

6-band  receiver  covers  80-6  meters,  using 
TV  turret  for  "front  end."  Mechanical  fil- 
ters provide  optimum  selectivity. 


filter  passband.  The  proper  pitch  control  set- 
ting may  be  determined  by  tuning  the  re- 
ceiver across  a carrier  while  adjusting  the  pitch 
control  so  that  a beat  note  on  only  one  side 
of  zero  beat  is  heard.  After  marking  this  setting 
of  the  pitch  control,  again  turn  it  so  the  test  sig- 
nal on  only  the  other  side  of  zero  beat  is  heard. 
Mark  this  setting,  then  try  tuning  in  a single 
sideband  signal.  If  intelligible  speech  cannot 
be  heard,  shift  the  b-f-o  pitch  control  to  the 
other  setting.  Small  adjustments  of  the  pitch 
control  and  the  tuning  of  the  receiver  will 
produce  the  correct  voice  tones.  For  best  re- 
ception, the  audio  control  of  the  receiver  should 
be  fully  advanced,  and  only  enough  r-f  gain 
used  to  produce  a readable  signal.  Volume  is 
therefore  controlled  by  the  r-f  gain  control. 

For  optimum  c-w  reception,  the  0.5  kc.  me- 
chanical filter  may  be  used.  No  changes  need  be 
made  to  the  adapter  when  filters  are  changed. 


27-4  A High 

Performance 
Amateur  Band  Receiver 

This  receiver  was  designed  for  outstanding 
performance  on  the  80,  40,  20,  15,  10,  and  6 
meter  amateur  bands.  It  incorporates  many  up- 
to-date  features  such  as  mechanical  filters  for 
optimum  selectivity,  double  conversion  for 
freedom  from  images,  and  a novel  r-f  assembly 
for  mechanical  stability  and  good  operating  ef- 
ficiency on  the  higher  frequencies.  The  re- 
ceiver uses  12  tubes,  plus  a selenium  rectifier 
power  supply.  The  complete  r-f  assembly  is  de- 
signed about  a television  turret  tuner  having 
positions  for  12  separate  bands.  The  rugged 
construction  and  excellence  of  the  self-wiping 
contacts  of  the  TV  tuner  add  greatly  to  the 
overall  stability  of  the  receiver.  Complete  dial 
calibration  and  an  S-meter  provide  maximum 
mning  simplicity  and  logging  ease. 

Because  of  the  "snap-in”  coil  strips  of  the 
tuner  it  is  possible  to  provide  coils  to  tune  any 
narrow  segment  of  the  shortwave  range  be- 
tween 3.5  megacycles  and  75  megacycles.  For 
example,  coils  can  be  wound  to  cover  the  13, 
19,  21,  31,  and  49  meter  international  short- 
wave broadcast  bands  if  desired.  Panel  control 
of  the  selectivity  allows  instant  selection  of  the 
optimum  passband  for  either  voice  or  c-w  re- 
ception. The  receiver  is  entirely  self-contained 
except  for  an  external  speaker. 

Circuit  The  High  Performance  Ama- 

Oescription  teur  Band  Receiver  is  a double 
conversion  superhetrodyne  ( fig- 
ure 17).  A modified  TV  turret  tuner  is  em- 
ployed to  cover  six  amateur  bands.  The  turret 


VOLTAGE  A-VC 

REGULATOR  BRIDGE 


Figure  17 

BLOCK  DIAGRAM  OF  DE-LUXE  AMATEUR  COMMUNICATION  RECEIVER 


Amateur  Band  Receiver  541 


employs  a 6BQ7A  cascode  r-f  amplifier  and 
a 6J6  mixer-oscillator.  All  r-f  tuned  circuits 
are  changed  as  the  tuner  is  switched  from 
channel  to  channel.  The  intermediate  frequen- 
cy output  of  the  tuner  is  1695  kc.  A 6BA7 
pentagrid  mixer  tube  converts  the  first  inter- 
mediate frequency  to  the  second  i-f  channel, 
which  is  455  kc.  To  obtain  a high  order  of 
stability,  the  6BA7  conversion  oscillator  is  crys- 
tal controlled  at  2150  kc. 

Switch  Si  in  the  output  circuit  of  the  6BA7 
mixer  stage  permits  the  selection  of  one  of  the 
two  mechanical  filters.  In  this  particular  re- 
ceiver, a 6 kc.  and  a 3.1  kc.  filter  were  used 
for  "broad”  and  "sharp”  phone  reception. 
Three  stages  of  455  kc.  amplification  follow 
the  mechanical  filter.  The  gain  of  these  stages 
is  adjustable,  permitting  the  overall  receiver 
gain  to  be  set  at  a reasonable  value.  Too  much 
i-f  gain,  will  make  the  receiver  noisy  and  sub- 
ject to  overloading  effects  on  strong  signals. 
Too  little  i-f  gain  will  make  the  receiver  seem 
insensitive.  The  i-f  gain  is  controlled  by  poten- 
tiometer Ri  and  also  by  the  value  of  the  ca- 
thode resistors  in  the  second  and  third  i-f 
stages.  In  order  to  prevent  clipping  of  the  ex- 
cellent flat-top  response  of  the  6 kc.  mechan- 
ical filter,  small  coupling  capacitors  are  placed 
across  the  "top”  of  the  i-f  transformer  coils  to 
widen  the  "nose  response”  of  the  circuits.  The 
"skirt”  selectivity  of  the  i-f  strip  is  determined 
by  the  passband  of  the  mechanical  filter  and 


not  by  the  transformer  response.  To  conserve 
filament  power,  150  milliampere  6BJ6  tubes 
are  employed  in  the  i-f  amplifier. 

A 6AL5  dual  diode  is  used  as  a second 
detector  and  a-v-c  rectifier.  Automatic  volume 
control  voltage  is  applied  to  the  first  two  i-f 
stages,  and  a small  portion  of  the  a-v-c  voltage 
is  applied  to  the  6BQ7A  r-f  amplifier  and 
6BA7  mixer.  The  a-v-c  circuit  is  rendered  in- 
operative for  c-w  and  SSB  operation  by  seg- 
ment D of  switch  Si.  a simple  delayed  a-v-c 
system  is  used  to  allow  best  reception  of  weak 
signals-  Potentiometer  Ri  sets  the  delay  voltage 
applied  to  the  cathode  of  the  6AL5  a-v-c  diode. 

A second  6AL5  serves  as  a noise  limiter.  The 
threshold  value  of  limiting  is  set  by  control 
Ri,  and  the  noise  limiter  may  be  removed  from 
the  circuit  by  switch  S.i.  The  audio  output  of 
the  detector  passes  through  volume  control  R< 
and  is  further  amplified  by  one  triode  section 
of  a 12AT7  followed  by  a 6AK6  power  am- 
plifier stage.  The  second  section  of  the  12AT7 
functions  as  the  beat  frequency  oscillator.  The 
frequency  of  the  b-f-o  is  controlled  from  the 
panel  by  variable  capacitor  G. 

The  S-meter  circuit  consists  of  a single  va- 
cuum tube  voltmeter  using  a 12AU7  dual  tri- 
ode  in  a bridge  circuit.  The  bridge  is  balanced 
by  potentiometer  Re,  and  the  meter  sensitivity 
is  set  by  control  Ri.  Automatic  volume  control 
voltage  applied  to  one  grid  of  the  12AU7  up- 
sets the  balance  of  the  bridge  and  causes  a 


"TV-TUNER 


CONVERTED  TV-TUNER  FORMS  "FRONT-END"  OF  AMATEUR  BAND  RECEIVER 

Several  forms  of  replacement  TV-tuners  may  be  purchased  from  large  radio  stores.  Shown  above  is  typical 
"cascode  tuner"  to  be  used  in  this  receiver.  Check  your  tuner  for  circuitry  before  you  modify  it.  I-f  output 
frequency  is  unimportant,  since  i-f  components  of  tuner  ere  femoved.  Tuner  trimmers  permit  minor 
adjustments  to  be  made  to  alignment  of  r-f  stages. 

A — Oscillator-mixer  excitation  measuring  point  on  tuner 

CiA-B-C — 15  fififd.  per  section.  All-Star  Products  Co.,  Defiance,  Ohio.  Model  C3. 

Cs — 70  /xfifd.  ceramic.  Centralab  827B 

Cs — 15  iiij.  Bud  LC-1641.  (Mounted  atop  chassis  after  photographs  were  taken). 

Li-L: — See  Figure  27  ^or  coil  data 


NOTES  : 1 -«•  S££  F/LT£R  /NSTRUCr/ON  SHEET  FOR  CAPACITOR  SPECIFICATIONS. 
2.- ALL  RESISTORS  1-WATT  UNLESS  OTHERWISE  NOTED. 


ADJUST  SENS. 


Figure  18B 

SCHEMATIC  OF  AMATEUR  BAND  RECEIVER 


The  **froni-end**  circuit  is  shown  in  Figure  J8A.  The  tuning  capacitors  placed  across  the  input  and  output 
circuits  of  the  mechanical  filters  are  determined  by  the  choice  of  filter.  See  filter  specification  sheet. 

Type  P455C  filters  were  used  in  this  receiver. 

M — 0-7  d-c  milliammeter,  1-inch  diameter 

SiA-B — DPDT  rotary  switch.  Centralab  PA-2001  decks,  with  Centralab  PA-302  index  assembly 
StA-B-C-D—4  pole,  3 position  switch.  Centralab  PA-1013 

T — 7500  kc.  i-f  transformer.  J.  W.  Miller  Co.  512-W1.  Remove  turns  from  primary  and  secondary  to 
permit  resonance  at  7695  kc. 

T i-T g^-4SS  kc.  i-f  transformer.  J.  W.  Miller  Co.  12-C1,  or  equivalent 
Ts — 455  kc.  diode  transformer.  J.  W.  Miller  Co.  12-C2,  or  equivalent 
T u — 455  kc.  BFO  transformer.  J.  W.  Miller  Co.  1 12-C5,  or  equivalent 
Ts — 750-0-750  volts  at  100  ma.  UTC  S-51 
Ts — 70  K pri.,  4 ohm  sec.,  Stancor  A-3879 
Tt — 6.3  volts,  3 amp.,  Stancor  P-6466 


542  Receivers  and  Transceivers  THE  RADIO 


HANDBOOK 


Amateur  Band  Receiver  543 


reading  on  the  0-1  d-c  milliammeter  connected 
between  the  plates  of  the  12AU7  triodes. 

The  selenium  rectifier  power  supply  delivers 
140  volts  at  the  receiver  current  drain  of  65 
milliamperes.  The  low  plate  voltage  and  small 
power  consumption  of  the  receiver  result  in  a 
very  low  degree  of  thermal  drift.  If  desired, 
the  selenium  supply  may  be  replaced  with  a 
more  conventional  supply  using  a 5Y3-GT 
rectifier  tube.  Plate  voltage  to  the  receiver 
should  be  held  below  160  volts.  Plate  voltage 
to  the  6J6  oscillator  section  and  the  12AU7 
S-meter  tube  is  regulated  by  an  OB2  gas  reg- 
ulator. 

Plate  voltage  is  removed  from  the  r-f  and 
audio  stages  when  the  receiver  is  in  "standby” 
condition.  All  oscillators  are  left  running  con- 
tinuously to  reduce  initial  warmup  drift. 

Receiver  The  receiver  is  built  upon  a plated 

Layout  steel  chassis  measuring  8"  x 15”  x 

2V2".  Panel  height  is  8".  To 
achieve  maximum  rigidity,  the  panel  is  braced 
to  the  chassis  with  two  steel  angle  brackets 
visible  in  figure  19.  The  tuning  dial  is  cen- 
tered on  the  panel,  and  the  three  gang  tuning 
capacitor  is  behind  the  dial,  on  the  center-line 
of  the  chassis.  Direaly  to  the  left  of  the  tuning 
capacitor  is  a 3V^”  x 5"  cut-out  in  the  chassis. 
The  TV  tuner  is  placed  in  this  cut-out,  held 
in  place  by  two  3"  angle  brackets  cut  from 
aluminum  stock.  Height  of  the  tuner  above 
the  chassis  is  determined  by  the  shaft  of  the 
tuner  which  must  pass  through  the  bushing 
drilled  in  the  lower  left  corner  of  the  dial 
escutcheon.  Likewise,  the  placement  of  the 


three  gang  tuning  capacitor  Ci  is  determined  by 
the  shaft  height  of  the  main  dial  bushing.  G 
is  therefore  mounted  above  the  chassis  by 
means  of  long  6-32  bolts  and  steel  spacers. 

Placement  of  the  major  components  may  be 
seen  in  figures  19  and  20.  The  two  mechanical 
filters  are  mounted  to  the  left  of  the  tuner, 
directly  behind  the  6BA7  mixer  tube  and  the 
1695  kc.  i-f  transformer,  Ti.  The  filter  selec- 
tion switch  Si  is  mounted  below  chassis  so  that 
the  switch  segments  fall  over  the  filter  ter- 
minals. This  switch  is  made  up  from  a switch 
kit,  and  is  mounted  in  place  by  means  of  two 
brackets  cut  from  soft  aluminum.  One  bracket 
holds  the  front  panel  bushing  of  the  switch, 
and  the  other  is  placed  just  forward  of  section 
SiB  and  holds  the  two  fixed  shafts  of  the 
switch.  A hole  is  cut  in  the  center  of  the 
bracket  to  clear  the  rotary  shaft  of  the  switch. 
This  bracket  also  serves  as  a shield  isolating 
the  output  circuits  of  the  mechanical  filter  from 
the  input  circuits.  The  particular  switch  used 
is  highly  recommended,  and  it  is  suggested  that 
no  substitution  be  made.  It  is  imperative  that 
the  drive  shaft  of  the  switch  be  metal,  and  that 
it  is  securely  grounded.  Some  switches  have 
fibre  drive  shafts,  or  metal  shafts  held  in  po- 
sition with  a cotter  pin.  Fibre  drive  shafts 
permit  signal  leakage  along  the  shaft,  deterior- 
ating filter  performance,  and  the  cotter  pin- 
type  switch  does  not  insure  the  shaft  is  secure- 
ly grounded  which  is  mandatory,  when  the 
metal  shaft-type  switch  is  used. 

The  three  intermediate  frequency  stages  are 
spaced  along  the  back  of  the  chassis.  At  the 
right  rear  are  the  6AK6  audio  stage,  the  volt- 


Figure  19 
TOP  VIEW  OF 
AMATEUR 
COMMUNICATION 
RECEIVER 

Mechanical  filiefs  are  at 
left  with  three  stage  /-/ 
amplifier  along  rear  of 
chassis.  The  TV  tuner  it 
mounted  in  chassis  cui~ 
out  to  the  left  of  main 
tuning  capacitor.  Power 
supply  and  audio  section 
occupy  right-hand  sec- 
tion of  steel  chassis.  Six 
channels  of  tuner  are 
used  to  cover  the  ama- 
teur bands  between  80- 
and  6-meters.  Six  spare 
channels  may  be  used  fo 

cover  shortwave  broad- 

cost  bands,  or  portions 
of  100  Me*  region. 


544  Receivers  and  Transceivers 


THE  RADIO 


age  regulator  tube,  and  the  12AU7  S-meter 
tube.  In  front  of  the  last  i-f  transformer  can 
are  the  two  6AL5  tubes  and  potentiometers 
Re  and  Re. 

At  the  right  end  of  the  chassis  is  the  power 
transformer,  and  between  it  and  the  panel  is 
the  b-f-o  transformer  and  the  12AT7  au- 
dio/b-f-o  tube.  Between  the  power  transformer 
and  the  three  gang  runing  capacitor  are  the 
power  supply  filter  capacitor  and  the  auxiliary 
filament  transformer.  The  i-f  gain  control  po- 
tentiometer Ri  is  placed  between  the  filament 
transformer  and  the  main  tuning  capacitor. 

Placement  of  the  smaller  components  below 
the  chassis  may  be  seen  in  figure  20.  The 
TV  turret  and  mechanical  filter  switch  are  at 
the  right  of  the  chassis,  with  the  audio  output 
transformer  mounted  to  the  back  wall  of  the 
chassis  near  the  turret.  The  filter  choke  for  the 
power  supply  is  mounted  to  the  left  side  wall 
of  the  chassis,  with  the  two  selenium  rectifiers 
in  front  of  it.  The  10  watt  dropping  resistor 
for  the  regulator  tube  is  mounted  to  the  chas- 
sis beween  the  filter  choke  and  the  rear  tube 
sockets. 

Receiver  The  receiver  should  be  built 

Construction  in  two  parts  to  simplify  con- 
struction. The  power  supply, 
audio  and  455  kc.  i-f  sections  ate  built  first, 
and  the  receiver  is  placed  in  operating  condi- 
tion up  to  the  1695  kc.  i-f  channel.  The  TV 
tuner  is  then  modified,  using  the  rest  of  the 
receiver  for  alignment  and  adjustment  tests. 

All  major  components  should  be  marked  in 
position  on  the  paper  wrapper  of  the  chassis 


which  is  then  drilled  and  punched.  All  major 
components  with  the  exception  of  the  TV 
turret  are  mounted  in  place.  The  ground  con- 
nections at  each  tube  socket  are  made  first, 
then  the  filament  leads  are  wired  in  position. 
Use  #14  insulated  wire  for  the  filament  leads, 
as  the  current  drain  is  quite  high.  The  power 
supply  should  be  wired  first  and  tested.  It 
should  deliver  approximately  140  volts  across 
a 2000  ohm,  10  watt  resistor  used  as  a rem- 
porary  dummy  load.  When  the  power  supply 
section  is  completed,  the  audio  stages,  b-f-o, 
S-meter,  second  detector,  and  noise  limiter 
stages  can  be  wired.  Operation  of  the  audio 
system  may  be  checked  by  injecting  a small 
audio  signal  on  pin  #2  of  the  12AT7  stage 
and  checking  the  volume  of  the  tone  in  the 
loud  speaker. 

The  next  step  is  to  wire  the  three  i-f  am- 
plifiers and  the  a-v-c  circuit.  Five  terminal 
phenolic  tie-point  strips  are  mounted  in  front 
of  each  i-f  tube  socket.  The  a-v-c,  cathode  bias, 
and  screen  dropping  resistors  are  mounted  be- 
tween the  socket  pins  and  the  terminals  of  the 
strip.  The  ceramic  bypass  capacitors  are  sol- 
dered directly  between  socket  pins  with  the 
shortest  possible  leads.  All  interconnecting 
wiring  is  done  between  the  terminal  strips. 
After  the  i-f  strip  has  been  wired  up  to  the 
arm  of  switch  Si,  segment  B,  the  amplifier 
may  be  tested.  Pin  # 1 of  the  first  6BJ6  i-f 
tube  is  temporarily  returned  to  the  a-v-c  bus 
by  placing  a 100  K resistor  across  the  output 
pins  of  the  mechanical  filter  socket.  Switch  Si 
is  adjusted  to  place  the  resistor  in  the  circuit. 


Figure  20 
UNDER-CHASSIS 
VIEW  OF  AMATEUR 
BAND  RECEIVER 

To  the  left  of  the  chassis 
are  the  selenium  recti- 
fiers and  the  filter  choke. 
At  the  extreme  right  is 
the  mechanical  filter  se- 
lector switch  mounted 
on  two  aluminum  brack- 
ets. The  controls  along 
the  panel  are  (left  to 
right);  BFO,  AUDIO 
GAIH,  MAN—AM—CW, 
tuning  control,  TV  tur- 
ret switch,  R-F  GAIN, 
and  filter  selector  switch. 
Output  transformer  is 
mounted  to  back  of 
chassis. 


HANDBOOK 


Amateur  Band  Receiver  545 


A signal  generator  having  a tone  modulated 
455  kc.  signal  is  attached  to  the  input  grid  of 
the  6BJ6  tube,  and  an  a-c  voltmeter  is  con- 
nected across  the  speaker  terminals.  To  insure 
that  the  i-f  system  is  aligned  at  the  center  of 
the  filter  passband  it  is  imperative  that  the 
alignment  frequency  be  very  close  to  455.0 
kc.  The  use  of  a BC-221  or  equivalent  signal 
generator  is  recommended.  As  the  transform- 
ers are  brought  into  alignment  the  input  signal 
to  the  i-f  strip  should  be  decreased.  A vacuum 
tube  voltmeter  may  be  employed  to  measure 
the  negative  voltage  developed  on  the  a-v-c  bus. 
A maximum  voltage  of  about  8 volts  is  de- 
veloped with  a strong  input  signal  to  the  i-f 
strip.  Overall  gain  is  controlled  by  the  setting 
of  Ri  when  switch  Ss  is  placed  in  the  "A-M” 
position. 

After  i-f  alignment  has  been  completed,  the 
100  K input  resistor  should  be  removed  from 
the  filter  socket  and  the  6BA7  mixer  stage 
wired.  The  capacitors  connected  between  pins 
1 and  2 and  ground  of  the  6BA7  socket  de- 
termine the  level  of  excitation  to  the  2150  kc. 
crystal.  If  difficulty  is  encountered  in  obtain- 
ing consistent  oscillation,  the  value  of  one  or 
both  of  these  capacitors  may  be  altered  until 
reliable  oscillation  is  obtained.  Operation  of 
the  crystal  oscillator  may  be  monitored  with  a 
nearby  receiver  tuned  to  the  crystal  frequency. 
When  the  oscillator  is  working  properly,  a 
1695  kc.  tone  modulated  signal  should  be 


coupled  through  a 0.01  gfd.  blocking  capaci- 
tor to  the  "hot”  end  of  the  primary  winding 
of  transformer  T.  This  transformer  is  adjusted 
for  maximum  output  signal.  The  receiver  is 
now  ready  for  tuner  installation. 

Tuner  The  tuner  used  in  this  re- 

Modification  ceiver  is  a Standard  Coil  Co. 

replacement  type  unit  using 
a 6BQ7A  cascode  r-f  amplifier  and  a 6j6  mix- 
er stage.  Several  different  variations  of  this 
tuner  can  be  purchased  ar  nominal  prices 
through  large  wholesale  outlet  stores.  The 
tuner  has  snap-in  coil  strips  for  12  TV  chan- 
nels, and  may  have  either  a 21  Me.  or  42  Me. 
i-f  output  range.  Several  simple  changes  must 
be  made  to  the  tuner  for  use  in  this  receiver. 
The  i-f  output  coil  and  the  cascode  neutralizing 
circuit  must  be  removed.  The  besr  thing  to  do 
is  to  remove  all  the  clip-in  coils  and  trace  out 
the  tuner  circuit,  and  then  compare  it  with  the 
circuit  shown  in  figure  1 8 A.  Cathode  bias  must 
be  added  to  the  first  section  of  the  cascode  and 
several  of  the  resistors  in  the  tuner  must  be 
changed.  Rewire  the  tuner  according  to  the 
circuit  of  figure  18 A.  Note  that  regulated  volt- 
age is  applied  to  the  oscillator  section  of  the 
6J6  and  that  a-v-c  voltage  is  applied  to  the 
first  section  of  the  6BQ7A.  The  main  tuning 
capacitor  ClA-B-C  is  connected  to  the  appro- 
priate points  in  the  tuner  by  three  short  leads. 
A M-inch  coil  is  drilled  in  the  rear  of  the 


F IGURE  21  - COIL  TABLE  FOR  AMATEUR  BAND  RECEIVER 


BAND 

COIL 

P 

WINDING  AND  WIRE  SIZE 

SPACING  ^ 

SHUNT  NATURAL 
APACITY  RESONANCE 

BAND 

COIL 

WINDING  AND  WIRE  SIZE 

SHUNT  NATURAL 
SPACING  CAPACITY  RESONANCE 

6M 

ANTi  3T.  N«  22  E. 

C-W 

— 

Li 

ANT:  7 T.  N*  22  E. 

C-W 

- 

GRIQiSr.  N*  18  E. 

WIRE  DIA. 

56  JUUFD. 

GRID:  13  T.  N*  22  E. 

Lz 

4«.3- 

52.3 

MC. 

PLATE,  R-F;  5T.  N^IBE. 

WIREDIA. 

62UUFO 

PLATE,  R-F  s |3T.  N*  22  E. 

GRID,  OSC. : ST.  N*  16  E. 

WIRE  DIA. 

62JUUFD. 

GR 1 D,  OSC. : 1 2 T.  N*  22  E. 

PLATE,  OSC.'  5T.  N»16  6. 

C-W 

— 

— 

C-W 

- 

— 

10M 

Li 

ANT.:  JOT.  N«  22  E. 

C-W 

— 

— 

B 

ANT.!  15  T.  N»  30  E. 

C-W 

- 

— 

GRIDS  10T.  N*  22  E. 

WIRE  DIA. 

75JJUF0. 

GRID:  39T.  N*  30  E. 

C-W 

7.9  MC. 

Lz 

26.3- 

26.0 

Me. 

PLATE,  R-F  : 10T.  N*22E. 

WIRE  DIA. 

BZUVfO. 

40M 

1 

L2 

6.7- 

9.1 

MC, 

PLATE,  R-F:  36 T.  N*30  6. 

7.6  MC. 

GRID,  OSC.  s 10 T.  N*  22  E. 

WIRE  DIA. 

eOJJJLIFO. 

E^Q^SH 

GRID.  OSC.:  25 T.  N*30E. 

C-W 

10  MC. 

PLATE,  OSC. ; 9 T.  N*  30  E. 

C-W 

— 

— 

PLATE,  OSC.  ••  12T.  N«30  6. 

C-W 

— 

— 

B 

ANT:  9 T.  N«  22  E. 

C-W 

— 

— 

B 

ANT  : 10  T.  N*  30  E. 

C-W 

— 

— 

GRID:  9T.  N«  22  E. 

CPID-  "LOOPSTICH"  B-C  COIL.  RE- 
GRID-  /I40YE7VRNS  TO  RESONANCE. 

BBElaBI 

PI 

PLATE.  R-F:  JOT.  N*22  E. 

n 

PLATE,  R-F:  SEE  Li  ABOVE 

12JJJJFD 

S.7  MC. 

GRID, OSC.;  10T.  N*22E. 

21.0  MC. 

GRID, OSC.:  SEE  Li  ABOVE 

PLATE, OSC. : 10T.  N*30E. 

IIBI 

— 

- 

- 

NOTE  1 -GRID  WINDING  OF  OSCILLATOR  COIL  LZ  IS  PLACED  NEXT  TO  OSCILLATOR  PLATE  WINDING.  R-F  PLATE  WINDING 
IS  PLACED  ABOUT  1/2~INCH  AWAY  FROM  OSCILLATOR-GRID  WINDING. 


Z-ANTENNA  COIL  SPACED  ABOUT  1/IB-INCH  FROM  GRID  COIL. 

Z-10  K RESISTOR  PLACED  ACROSS  tO  METER  R-F  GRID  COIL. 

4 - RESONANT  FREQUENCY  MEASURED  WITH  SHUNT  CAPACITOR  ACROSS  COIL  . 


546  Receivers  and  Transceivers 


THE  RADIO 


tuner  case  to  permit  the  coaxial  lead  from  the 
antenna  receptacle  to  enter  the  tuner.  The 
tuner  is  now  ready  to  be  mounted  in  the  re- 
ceiver. 

Tuner  Coil  The  tuner  coils  should  be 

Modification  removed  from  their  snap-in 
holders.  Notice  that  the  r-f 
coils  have  five  terminals  and  that  the  mixer- 
oscillator  coils  have  six  terminals.  All  coil 
forms  and  windings  are  removed  from  the 
clip-in  assemblies  and  may  be  discarded.  The 
new  coils  for  all  bands  are  to  be  wound  on 
lengths  of  54 -inch  diameter  polystyrene  rod, 
cut  to  fit  into  the  assembly  clamps.  For  this 
particular  tuner,  the  r-f  coil  forms  are  1 54” 
long  and  the  oscillator  mixer  coil  forms  are 
1 13/16"  long.  The  ends  of  the  forms  are 
tapered  slightly  with  a file  so  that  they  fit 
into  the  slotted  ends  of  the  assembly  clamp. 
After  each  set  of  coils  is  completed  the  fre- 
quency band  notation  is  marked  on  the  unit 
for  future  identification. 

Complete  coil  data  is  given  in  figure  21 
and  a complete  set  of  coils  is  shown  in  figure 
22.  Start  with  the  ten  meter  coils,  as  they  have 
few  turns  and  are  easy  to  wind.  Make  the  r-f 
stage  coil  first.  The  windings  may  be  tem- 


porarily held  in  place  with  ^ small  piece  of 
cellophane  tape  until  the  coil  form  is  cemented 
in  the  assembly  clamp  with  Duco  cement. 
The  coil  leads  are  then  cleaned  and  soldered 
to  the  corresponding  terminals.  After  the  shunt 
capacity  is  added,  the  natural  resonance  of  the 
winding  should  be  checked  against  the  value 
given  in  figure  21.  The  coil  should  be  sus- 
pended in  the  clear  and  checked  with  a grid- 
-dip  oscillator.  Turn  spacing  may  be  varied  to 
produce  the  correa  resonance  point.  The  mix- 
er-oscillator coil  is  wound  next.  The  two  tuned 
windings  are  placed  at  the  ends  of  the  form 
and  the  oscillator  plate  winding  is  placed  be- 
tween them,  positioned  close  to  the  grid  wind- 
ing as  illustrated  in  figure  22.  The  coil  wind- 
ings may  be  held  in  position  with  a temporary 
piece  of  cellophane  tape.  This  coil  is  adjusted 
in  the  same  manner  as  the  r-f  coil.  When  com- 
pleted, the  mixer-oscillator  coil  may  be  snapped 
into  the  tuner  and  the  receiver  turned  on.  The 
oscillator  operates  on  the  low  frequency  side 
of  the  signal  for  10  meter  reception,  cover- 
ing the  range  of  26.3  Me.  to  28.0  Me.  as  the 
receiver  is  tuned  from  27,995  kc.  to  29.7  Me. 
Listen  for  the  oscillator  in  a nearby  receiver. 
The  frequency  range  of  the  oscillator  can  be 
set  by  varying  the  spacing  of  the  turns  of  the 
6J6  oscillator  coil,  and  also  by  adjustment  of 
the  oscillator  trimmer  capacitor  Cz  in  the  tuner. 
When  the  correct  frequency  range  is  covered 
the  winding  may  be  permanently  held  in  place 
with  a small  application  of  Duco  cement.  Check 
the  frequency  range  once  again  after  you  ap- 
ply the  cement  as  the  frequency  shifts  slightly 
lower  as  the  cement  is  applied.  Make  your 
final  correction  before  the  cement  dries.  The 
other  coil  windings  are  relatively  non-critical 
and  require  no  adjustments  after  they  are 
completed. 

Once  the  oscillator  tuning  range  has  been 
set,  the  signal  generator  may  be  used  to  cali- 
brate the  main  dial  of  the  receiver.  If  care  is 
used  in  this  process,  the  dial  calibrations  can 
be  read  to  an  accuracy  of  five  kilocycles  of 
less.  The  oscillator  injection  voltage  may  be 
checked  at  terminal  A on  the  tuner  with  a 
v-t-v-m.  The  injection  range  should  lie  between 
two  to  three  volts  on  each  band,  and  may  be 
controlled  by  varying  the  spacing  between  the 


Figure  22 

COIL  SET  FOR  AMATEUR 
BAND  RECEIVER 

Tuner  ceils  are  wound  on  poly- 

styrene rods.  R-f  coils  are  at  left,  with 
mixer-oscillator  coils  at  right.  80  meter 
coils  are  at  top  of  photo,  with  6 meter 
coils  at  bottom. 


1 


HANDBOOK 


"Handie-Talkie"  547 


oscillator  grid  winding  and  the  untuned  plate 
winding.  These  adjustments  may  be  made  with 
the  coils  in  the  receiver  by  removing  the  bot- 
tom and  side  plates  of  the  tuner. 

The  conversion  oscillator  operates  on  the 
high  frequency  side  of  the  signal  on  the  15, 
20,  40,  and  80  meter  bands,  as  tabulated  in 
figure  21.  The  80  coils  are  made  from  "loop- 
stick”-type  broadcast  coils.  Turns  are  removed 
from  these  coils  until  the  natural  resonance 
falls  at  the  desired  frequency  as  shown  in  the 
table.  A small  value  of  capacitance  is  used 
across  these  coils  since  the  percentage-tuning 
range  is  quite  high. 

Receiver  After  the  first  set  of  coils  has 

Adjustment  been  completed,  the  receiver 

may  be  adjusted  for  optimum 
results.  A balance  between  r-f  gain  and  i-f 
gain  must  be  achieved.  Under  normal  operat- 
ing conditions,  the  residual  noise  picked  up 
by  the  antenna  system  should  mask  out  the  in- 
ternal noise  level  of  the  receiver.  If  the  an- 
tenna is  removed  from  the  receiver  and  the 
internal  noise  level  does  not  drop  appreciably 
it  is  usually  a sign  that  the  i-f  gain  of  the 
receiver  is  too  high  and  the  r-f  gain  is  too 
low.  The  value  of  i-f  gain  may  be  controlled 
by  the  setting  of  Ri.  Normally,  there  is  no 
instability  or  regeneration  in  the  i-f  strip.  The 
stability,  however,  is  dependent  to  some  extent 
upon  the  wiring  technique  you  use;  it  may 
be  necessary  to  retard  the  i-f  gain  level  to 
prevent  instability.  Receiver  i-f  gain  may  be 
further  decreased  by  increasing  the  value  of 
the  cathode  resistor  of  the  first  6BJ6  stage. 
I-f  gain  may  be  increased  by  decreasing  the 
value  of  the  cathode  resistors  of  the  second 
two  i-f  stages. 

The  a-v-c  bias  delay  potentiometer  Rj  should 
be  set  to  provide  about  one  volt  of  bias  on 
pin  # 1 of  the  6AL5  a-v-c  rectifier.  This  will 
retard  a-v-c  action  on  weak  signals  and  im- 
prove the  signal-to-noise  ratio  of  threshold 
signals. 

The  level  of  beat  oscillator  injection  can  be 
set  by  varying  the  setting  of  coupling  capaci- 
tor C<.  Sufficient  injection  should  be  used  to 
provide  good  reception  of  single  sideband  sig- 
nals. The  injection  level  is  not  critical  for 
c-w  operation.  Finally,  the  b-f-o  capacitor  G 
should  be  set  at  mid-scale  and  the  slug  of 
transformer  T4  adjusted  until  the  b-f-o  signal 
is  tuned  exactly  to  455  kc. 

The  response  of  the  S-meter  is  controlled 

by  the  time  constant  in  the  grid  of  the  12AU7 

meter  tube.  If  the  grid  capacitor  in  this  circuit 
is  increased  in  value  from  .05  gfd.  to  1.0  /afd. 


the  S-meter  will  respond  much  more  slowly 
to  changes  in  signal  strength.  The  value  of 
this  capacitor  can  be  altered  to  suit  the  user’s 
taste. 

Finally,  various  settings  of  the  noise  limiter 
threshold  control  Ri  should  be  tried  to  find 
the  best  clipping  level  for  various  types  of 
pulse-type  interference. 

Receiver  Receiver  tuning  is  extremely 

Operation  smooth,  and  the  dial  calibration 

remains  as  good  as  the  original 
calibration.  The  r-f  input  circuits  are  designed 
for  50-100  ohm  unbalanced  transmission  line 
on  all  amateur  bands,  and  r-f  stage  resonance 
may  be  easily  attained  by  means  of  the  an- 
tenna trimming  capacitor  located  within  the 
set.  For  phone  reception,  switch  82  is  set  to 
A-M  and  the  audio  level  is  controlled  by  the 
audio  potentiometer  R4.  For  c-w  or  sideband 
reception  82  is  set  to  C-W  and  audio  control 
Ri  is  fully  advanced.  Receiver  gain  is  now 
controlled  by  the  R-F  gain  potentiometer  Ri. 
The  B-F-O  pitch  control  Cs  is  adjusted  to  place 
the  oscillator  on  the  edge  of  the  filter  pass- 
band  for  "single  signal”  reception  of  c-w  sig- 
nals or  single  sideband  reception  of  88B  trans- 
missions. The  8-meter  is  operative  on  the 
"A-M”  position  of  82;  8-meter  sensitivity  is 
controlled  by  potentiometer  Rt. 

27-5  A "Handle  - 

Talkie"  for  144  Me. 

The  144  Me.  band  is  the  ideal  frequency 
range  for  short-haul,  portable  equipment.  Little 
power  is  required  for  local  coverage,  and  the 
antenna  size  is  moderate.  Then  too,  this  is  just 
about  the  highest  frequency  at  which  filament 
type  battery  tubes  will  operate. 

8hown  in  the  accompanying  illustrations  is 
a battery  powered,  crystal  controlled  "handie- 
talkie”  designed  for  operation  in  the  2 meter 
region.  It  is  especially  well  suited  for  Civil 
Defense  work  or  emergency  service.  This  unit 
was  designed  after  the  AN/URC-4  air-sea  res- 
cue transmitter,  and  is  completely  self-con- 
tained, including  batteries. 

Circuit  The  circuit  of  the  "handie- 

Description  talkie”  is  shown  in  figure  24. 

8ix  quick  heating  filament  type 
tubes  are  used.  Two  mbes  are  used  in  the 
transmitter  portion  of  the  transceiver,  two  in 
the  receiver  portion,  and  the  remaining  two 
in  the  common  audio  section. 

The  transmitting  portion  of  the  unit  em- 
ploys a 3A5  dual  triode  and  a 3A4  pentode. 


548  Receivers  and  Transceivers 


THE  RADIO 


The  first  triode  section  of  the  3A5  serves  as 
an  overtone  crystal  oscillator,  delivering  24 
Me.  energy  from  an  8 Me.  crystal.  The  crystal 
oscillates  in  series  mode  and  sufficient  regen- 
eration is  introduced  in  the  circuit  by  return- 
ing the  crystal  circuit  to  a tap  on  the  oscillator 
plate  coil.  The  oscillator  may  be  adjusted  by 
tuning  the  slug  of  coil  L4.  The  24  Me.  plate 
circuit  of  the  oscillator  stage  is  capacitively 
coupled  to  a 3A4  pentode  triplet  stage  whose 
plate  circuit  is  tuned  to  72  Me.  This  stage 
is  resonated  by  means  of  the  slug  of  coil  Li. 
The  3A4,  in  turn,  is  capacitively  coupled  to 
the  second  section  of  the  original  3A5  which 


Figure  23 

BATTERY  POWERED  "HANDIE- 
TALKIE"  FOR  2 METERS 


Completely  self-contained  transceiver  em- 
ploys crystal  controlled  transmitter  for 
greatest  stability.  Send-receive  switch  is  at 
top  right  with  tuning  control  of  receiver 
directly  below. 


serves  as  a doubler  from  72  Me.  to  144  Me. 
Tuning  of  the  plate  circuit  is  accomplished 
by  the  slug  of  coil  Lg.  Inductive  coupling  is 
used  between  Lg  and  the  quarter-wave  whip 
antenna. 

The  receiver  employs  a 5678  sub-miniature 
pentode  r-f  stage  to  provide  a better  signal-to- 
noise  ratio  and  also  to  reduce  spurious  radia- 
tion from  tbe  super-regenerative  detector.  At 
a distance  of  fifty  feet  from  the  transceiver, 
the  detector  radiation  is  inaudible  to  a sen- 
sitive receiver.  In  the  interest  of  economy,  the 
5678  can  be  replaced  with  a 959  "acorn”  tube 
if  miniaturization  of  the  unit  is  not  desired. 
The  super-regenerative  detector  uses  a special 
quench  circuit  which  provides  smooth  opera- 
tion yet  allows  the  usual  variable  quench  con- 
trol to  be  eliminated.  Values  shown  in  the 
schematic  are  optimum,  and  care  should  be 
taken  to  keep  all  leads  short.  Capacitor  Ci  is 
the  receiver  tuning  control  and  has  a range 
of  140  Me.  to  150  Me.  If  it  is  desired  to  re- 
duce the  tuning  range,  one  or  two  plates  can 
be  removed  from  Ci  and  coil  Ls  readjusted 
accordingly.  The  5676  can  be  replaced  by  a 
957  "acorn”  tube  in  the  interest  of  economy- 

Audio  output  from  the  super-regenerative 
detector  is  passed  through  a simple  R-C  filter 
to  suppress  the  quench  frequencies.  The  first 
audio  stage  is  a 1S5  pentode  using  contact  po- 
tential grid  bias.  Interstage  coupling  is  ac- 
complished by  a Centrdab  Couplate  printed 
circuit  incorporating  the  six  components  re- 
quired at  this  juncture.  If  this  item  is  not 
available,  the  values  of  resistance  and  capaci- 
tance shown  can  be  substituted.  The  output 
stage  consists  of  a 3Q4  power  pentode  driving 
a 214-inch  speaker  through  the  output  trans- 
former T2.  For  reception  only  one-half  of  the 
3Q4  filament  is  used  to  conserve  "A”  and  "B” 
battery  drain.  During  transmit  periods  the 
extra  audio  is  needed  and  the  other  half  of 
the  filament  is  energized. 

The  "send-receive”  switch  SIA-B-C-D-E  is 
a miniature  ceramic  rotary  switch.  Section  A 
switches  the  antenna  from  the  receiver  to  the 
transmitter.  Secion  B energizes  the  filaments 
of  the  receiver  r-f  tubes  on  the  "receive”  po- 
sition, and  energizes  the  microphone  and  the 
second  section  of  the  3Q4  filament  on  the 
"transmit”  position.  Section  C turns  off  all 
filaments  in  the  "off”  position,  and  section 
D opens  the  67.5  volt  battery  circuit  in  the 
"off”  position.  Section  E connects  the  loud- 
speaker to  the  audio  system  in  the  "receive” 
position  of  switch  Si.  The  extreme  counter- 
clockwise position  of  Si  is  the  "off”  position. 


HANDBOOK 


"Handie-Talkie"  549 


Under  normal  operating  conditions  the 
transceiver  is  held  like  a telephone  and  a short 
whip  antenna  is  used.  The  operating  range 
may  be  extended  by  removing  the  whip  and 
using  a beam,  connected  to  the  unit  by  a 
length  of  52-ohm  coaxial  transmission  line. 

Three  separate  battery  voltages  are  required 
by  the  "handie-talkie.”  IVi-volts  is  required  for 
the  filament  circuit,  4.5  volts  is  required  for 
grid  bias  of  the  3Q4  audio  stage,  and  67-5 
volts  is  required  for  the  plate  voltage.  All 
these  voltages  may  be  obtained  from  miniature 
batteries  placed  within  the  case  of  the  trans- 
ceiver. 

Transceiver  No  hard  and  fast  assembly  pro- 
Layout  cedure  is  given  for  this  unit, 

since  each  builder  will  want  to 
custom-fit  the  circuit  into  his  particular  case. 
The  unit  shown  in  the  photographs  was  built 


within  a steel  box  measuring  9”  long  by  2%" 
wide  by  lYs"  deep.  The  width  and  depth  di- 
mensions were  determined  by  the  size  of  the 
battery  pack.  The  "front”  of  the  cabinet  is 
one  piece  with  flanged  edges.  The  microphone 
and  speaker  are  attached  to  this  portion  of 
the  box.  The  transceiver  chassis  is  custom  de- 
signed to  fit  the  available  components  and  is 
held  to  the  front  of  the  box  by  two  6-32 
bolts  through  the  rear  of  the  speaker  frame. 
The  rear  and  sides  of  the  cabinet  serve  merely 
as  a dust  shield  as  no  components  are  fastened 
to  them,  except  for  the  coaxial  antenna  recep- 
tacle. The  portion  of  the  cabinet  directly  over 
the  batteries  is  separately  removable  to  permit 
easy  access  to  the  battery  pack.  The  remaining 
portion  of  the  cabinet  is  affixed  permanently 
to  the  front  piece  by  means  of  sheet  metal 
screws.  The  crystal  projects  through  a cut-out 
in  this  piece,  and  there  are  three  /s-inch  holes 


SCHEMATIC,  "HANDIE-TALKIE"  FOR  2 METERS 

Coil  data  is  given  in  Figure  26. 

Ci — /?  iiiifd.  "butterfly".  Johnson  3M6M 

SiA-B-C-D-E — 5 po/e,  three  position  rotary  switch.  Centraiab  PA-20iS 
Ti — 100  ohms  to  grid.  Triad  A-1X 
Ti — 5 K pri.f  6.7  K and  4 ohm  sec.  Triad  M-4Z 
Fi — Telephone-type  carbon  microphone. 


550  Receivers  and  Transceivers 


THE  RADIO 


through  which  the  transmitter  tuning  slugs 
may  be  adjusted.  These  holes  are  closed  with 
snap-in  hole  plugs  after  adjustments  are  com- 
pleted. 

A general  view  of  the  transmitter  chassis 
may  be  seen  in  figures  25,  27,  and  28.  Notice 
that  a small  shield  passes  across  the  r-f  stage 
tube  socket,  and  pin  3 of  the  socket  is  ground- 
ed to  the  shield.  The  shield  isolates  coil  Li 
from  tuning  capacitor  G which  is  placed  di- 
rectly below  it.  The  two  audio  stages  and 
transformer  Ti  mount  at  the  end  of  the  chassis. 
The  tubes  mount  parallel  to  the  front  panel 
on  a small  aluminum  angle  bracket.  Directly 
below  Ti  is  the  5676  detector  tube,  also  placed 
in  a horizontal  position. 

The  crystal  socket,  the  transmitter  coils,  the 
two  r-f  tubes,  and  the  "send-receive"  switch 
Si  are  mounted  at  the  far  end  of  the  chassis. 
A small  shield  plate  is  placed  between  coils 
L<  and  Ls  to  reduce  interstage  feedback.  Output 
transformer  T2  is  located  between  the  3A4 
plate  coil  (Ls)  and  the  audio  tubes. 

The  r-f  stage  coil  Li  is  wound  upon  a slug- 
tuned  form  made  from  a concentric  trimmer 
capacitor.  The  outer  electrode  of  the  capacitor 
is  removed,  exposing  the  polystyrene  form 
upon  which  the  coil  may  be  wound.  The  inner 
slug  of  the  capacitor  serves  as  the  tuning  slug 
of  the  coil. 


FIGURE  26-coil  TABLE  FOR  "HANDIE-TALKIE-I 

Ll-S  TURNS  N*  16  E.  rfjrr) 

L2-2  TURNS  N*22  INSULATED  WIRE  COUPLED 
TO  GRID  END  Of  COIL  Ls. 

LS-STURNS  N"16  E..  3/6-INCH  DIAMETER, 

1/2 -INCH  LONG,  AIR  WOUND. 

L4-  20  TURNS  N*  26  E.,  1/2-INCH  DIAMETER, 

1/2-INCH  LONG.  TAP  9 TURNS  PROM  GRID  END. 
RESONATES  IN  CIRCUIT  AT  24  MC. 

LS-  7 TURNS  N*  16  E.,  3/8-INCH  DIAMETER, 

1/2-IN.  LONG.  RESONATES  IN  CIRCUIT  AT  72  MC. 

L6  - S TURNS  N*16  E..  3/8-  INCH  DIAMETER, 

1/2 -IN.  LONG.  RESONATES  IN  Cl  RCUIT  AT  144  MC- 


Transceiver  Wiring  techniques  are  extreme- 

Wiring  ]y  important  when  working  on 

VHF  equipment  that  must  be 
compacted  into  a small  space.  It  is  recommend- 
ed that  a "pencil-type”  soldering  iron  be  used, 
along  with  small  gage  wire  whose  insulation 
will  not  melt  if  it  is  overheated.  All  socket 
grounds  and  filament  wiring  are  completed 
first.  Coils  are  wound  and  put  into  position. 
After  the  coupling  capacitors  and  other  r-f 
components  are  mounted  in  place,  the  tubes 
are  placed  in  the  sockets  and  the  tuned  circuits 
are  adjusted  to  frequency.  It  may  be  necessary 
to  alter  the  number  of  turns  on  coils  Li,  Ls,  Ls, 
and  L«  due  to  slight  variances  in  circuit  com- 
ponents and  tube  inter-electrode  capacities. 


Figure  25 
VIEW  OF 

"WALKIE-TALKIE" 
CHASSIS  SHOWING 
MAJOR 
COMPONENTS 

Oblique  view  of  chassis 
looking  fowards  switch 
5i.  Output  coil  Lg  is  in 
foreground  with  3AS 
tube  behind  it.  To  right 
of  3A5  is  3A4  tripler 
stage.  Coils  Li  and  Ls  are 
behind  3A4.  To  right  of 
switch  Si  is  receiver  tun- 
ing control  Cl.  Shield 
separates  Ci  from  coll 
Li  directly  above  it.  R’f 
amplifier  tube  5687 
mounts  on  vertical  plate 
supporting  Li.  Shield  be- 
low  r-f  tube  isolates  it 
from  detector  compon- 
ents below  the  chassis. 
Audio  tubes  are  at  right, 
with  Ti. 


Figure  27 
UNDER-CHASSIS 
VIEW  OF 
WALKIE-TALKIE 
SHOWING  MAJOR 
COMPONENTS 

Chassis  of  iransceiyer  is 
specially  bent  to  fit  com- 
ponents in  smallest  possi- 
ble space.  Oscillator  coil 
Lt  is  at  left  separated 
from  tripler  coil  Ls  by 
small  shield.  Output  coil 
Le  is  to  rear.  5676  detec- 
tor tube  is  at  right  rear, 
hanked  by  20  tifd.  HIter 
capacitor.  Audio  tubes 
are  under  chassis  parti- 
tion at  right.  Quench  fil- 
ter capacitors  are  to 
right  of  coil  Ls. 


These  adjustments  may  be  done  quickly  and 
easily  with  the  aid  of  a grid  dip  oscillator. 

After  the  r-f  circuits  are  adjusted  the  rest  of 
the  transceiver  may  be  wired-  Short,  direct 
leads  should  be  used  in  the  r-f  circuits  and  all 
wires  should  be  routed  in  the  shortest  possible 
paths. 

Transceiver  After  all  wiring  has  been  check- 
Alignment  ed,  the  tubes  should  be  removed 
from  their  sockets  and  the  fila- 
ment batteries  connected  to  the  unit.  The  fila- 
ment voltage  at  each  tube  socket  should  be 
checked  in  all  three  operating  positions  of 
switch  Si.  Make  sure  the  second  filament  sec- 
tion of  the  3Q4  (pin  #7)  Is  energized  in  the 
"transmit”  position  of  Si.  The  bias  batteries 
and  B-plus  batteries  may  now  be  attached  to 
the  proper  leads.  Be  sure  of  the  correct  battery 
polarity  before  inserting  the  tubes  in  the  sock- 
ets. Test  the  "handie-talkie”  for  proper  recep- 
tion first.  The  slug  of  coil  Li  should  be  adjust- 


ed for  best  reception  after  the  transceiver  is 
assembled  and  the  whip  antenna  placed  In  the 
coaxial  receptacle.  However,  It  is  possible  to 
place  a short  piece  of  wire  in  the  antenna  re- 
ceptacle to  pick  up  a test  signal  while  the 
transceiver  is  dissembled  on  the  bench.  The 
coupling  between  Ls  and  Ls  should  be  adjusted 
for  best  super-regenerative  action  and  sensi- 
tivity . 

After  the  receiver  is  operating,  the  transmit- 
ter should  be  energized.  Tune  the  slug  of  coll 
L<  for  reliable  crystal  operation.  A grid  dip 
oscillator  held  in  the  proximity  of  Ls  will  serve 
as  a r-f  indicator.  Remove  the  plate  voltage 
from  the  grid  dip  oscillator  for  this  test.  Al- 
ternatively, oscillator  operation  may  be  moni- 
tored in  a nearby  receiver  tuned  to  the  24 
Me.  overtone  frequency  of  the  crystal.  The 
3A4  tripler  stage  may  now  be  tuned  for  maxi- 
mum output  on  72  Me.  as  indicated  on  the 
grid  dip  oscillator.  Finally,  the  grid  dip  oscilla- 
tor (or  wavemeter)  is  held  in  close  proximity 


Figure  28 
TRANSCEIVER 
CHASSIS  BOLTED 
TO  PANEL  OF 
WALKIE-TALKIE 
The  chassis  is  held  in 
place  by  bolts  passed 
through  frame  of  small 
speaker.  Modulation 
transformer  Tt  is  visible 
in  foreground  with  Cen- 
tralab  "Coupiate"  at  rear 
of  audio  tube  sockets. 
Batteries  occupy  space 
to  left  above  microphone. 
Crystal  socket  is  at  far 
right,  which  is  "top"  end 
of  transceiver. 


■■I* 


552  Receivers  and  Transceivers 


THE  RADIO 


to  the  144  Me.  doubler  plate  coil,  Le.  This 

coil  is  tuned  for  maximum  output  at  2 meters. 
After  this  preliminary  alignment  is  completed 
the  chassis  of  the  transceiver  may  be  placed  in 
the  case,  along  with  the  battery  power  pack. 

Finol  Experience  with  several  of  these 

Alignment  units  has  shown  that  maximum 
output  is  obtained  with  a whip 
antenna  slightly  longer  than  quarter  wave- 
length. The  transmitter  stages  should  be  re- 
tuned  for  maximum  field  strength  in  a nearby 
wavemeter  and  different  length  whip  antennas 
should  be  tried.  The  quarter  wavelength  whip 
is  nineteen  inches  long,  but  whip  lengths  up 
to  twenty-seven  inches  should  be  tried.  In 
general,  whip  lengths  of  twenty-four  inches  or 
so  seem  to  produce  best  results  with  this  little 
transceiver.  Input  to  the  final  stage  of  the 
3A5  is  approximately  /2-watt  under  optimum 
loading  conditions. 

The  operating  range  between  two  of  these 
units  varies  from  a thousand  feet  or  so  over 
obstructed  terrain  to  better  than  three  miles 
over  smooth  ground.  Communication  between 
the  "handie-talkie”  and  a base  station  equipped 
with  a sensitive  receiver  and  a high  gain  beam 
can  be  maintained  over  greater  range. 

27-6  Six  Meter 
Transceiver 
for  Home  or  Cor 

The  six  meter  band  has  become  extremely 
popular  within  the  past  few  years.  The  pin- 
nacle of  the  sunspot  cycle,  combined  with  the 
influx  of  new  amateurs  has  brought  life  to 
an  otherwise  neglected  amateur  band.  Most 
of  the  stations  operating  on  this  band  employ 


relatively  low  power  and  the  amateur  having 
modest  power  suffers  no  handicap  if  he  has  a 
good  antenna. 

The  growth  of  mobile  operation  on  the  six 
meter  band  has  been  rapid  and  the  need  has 
arisen  for  a compact  transceiver  that  will  work 
well  either  in  the  car  or  at  home.  The  unit 
described  in  this  section  has  been  designed  to 
meet  this  need. 

This  compact  transceiver  operates  in  the  50- 
54  Me.  range.  It  employs  a stable,  superhetero- 
dyne type  receiver  and  a 10  watt  transmitter. 
The  transmitter  may  be  either  crystal  con- 
trolled, or  driven  with  an  external  variable 
frequency  oscillator.  A 5 -watt  audio  system 
operates  from  a high  impedance  crystal  micro- 
phone and  provides  100%  modulation  of  the 
transmitter.  In  the  receiver  mode,  the  audio 
system  delivers  sufficient  power  to  drive  an 
external  speaker  well  above  the  noise  level 
of  the  automobile. 

Small  enough  to  fit  comfortably  under  the 
dash  of  today’s  car,  the  transceiver  delivers  a 
well  modulated  signal  at  a total  plate  power 
drain  of  100  milliamperes  at  250  volts  po- 
tential. Voltage  regulation  of  the  plate  and  fila- 
ment voltages  of  the  receiver  oscillator  en- 
sures the  maximum  stability  required  for  mo- 
bile work. 

Transceiver  A block  diagram  of  the  trans- 
Circuit  ceiver  is  shown  in  figure  30. 

Eleven  tubes  are  employed,  plus 
a filament  voltage  regulator.  Change-over 
from  receive  to  transmit  is  accomplished  by  a 
push-to-talk  circuit  that  operates  two  d-c  re- 
lays from  the  microphone  control  button. 

The  receiver  employs  a broad-band  r-f  stage 
and  two  i-f  stages.  The  r-f  stage  uses  a 6CB6 
television-type  pentode  with  pre-tuned  grid  and 
plate  circuits.  This  stage  is  capacity  coupled 
to  a second  6CB6  mixer  operating  with  grid 
injection.  The  variable  frequency  oscillator  is 
a third  6CB6  in  a Tri-tet  configuration.  The 
oscillator  circuit  (Ls-Cs-Cj)  tunes  the  17.0- 
17.7  Me.  range.  The  plate  circuit  of  this 


Figure  29 

DE-LUXE  SIX  METER  TRANSCEIVER 
FOR  HOME  OR  CAR 

Operating  from  a separate  power  package, 
this  compact  transceiver  is  we//  suited  for 
the  serious  six  meter  operator.  A 10  watt 
crystai  controlied  transmitter  and  super- 
heterodyne receiver  are  combined  with  a 
dual  purpose  audio  system  in  this  smatl 
cabinet.  Receiver  controis  are  at  ieft  and 
transmitter  controis  at  right.  Meter  serves 
as  5-meter  and  transmitter  tuning  indica- 
tor. External  VFO  may  be  used  with  trans- 
ceiver. 


HANDBOOK 


6-Meter  Transceiver  553 


stage  is  gang-tuned  to  the  oscillator  and  covers 
the  third  harmonic  range  of  51-53  Me.  The 
harmonic  voltage  is  capacitively  coupled  to  the 
input  grid  circuit  of  the  mixer  stage.  Series 
padding  capacitors  in  the  oscillator  and  tripler 
stages  permit  close  tracking  over  the  tuning 
range. 

An  i-f  channel  of  1 Me.  is  employed  in  this 
transceiver.  When  the  oscillator  is  considered 
to  be  on  the  high  frequency  side  of  the  received 
signal,  the  tuning  range  is  50-52  Me.  When 
the  oscillator  is  considered  to  be  on  the  low 
frequency  side  of  the  received  signal,  the 
tuning  range  is  52-54  Me.  The  broad  band  r-f 
circuits  of  the  receiver  "front  end”  cover  the 
complete  range  of  50-54  Me.  so  that  signals 
on  either  side  of  the  injection  frequency  may 
be  detected.  This  principle  is  sometimes  re- 
ferred to  as  super-imposition  tuning,  wherein 
the  receiver  is  tuned  to  two  frequencies  simul- 
taneously. For  example,  if  the  local  oscillator 
is  tuned  to  51  Me.,  simultaneous  reception  of 
signals  on  50  Me.  and  52  Me.  is  possible. 
Since  most  amateur  activity  takes  place  in  the 
lower  one  megacycle  of  the  six  meter  band, 
the  chance  of  two  signals  appearing  at  one 
point  of  the  dial  that  in  reality  are  two  mega- 
cycles apart  is  remote. 

Super-imposition  tuning  greatly  simplifies 
the  design  of  a VHF  receiver.  Double  conver- 
sion is  not  requited,  and  the  local  oscillator 
need  tune  only  a relatively  narrow  frequency 
range.  Saving,  in  parts  and  circuit  complexity, 
is  obvious. 


Two  stages  of  one  megacycle  i-f  amplifi- 
cation are  used  for  maximum  gain.  Five  i-f 
transformers  are  used  to  narrow  the  "skirt" 
response  of  the  i-f  strip.  A 12AL5  serves  as 
the  detector,  a-v-c  rectifier,  and  automatic 
noise  limiter.  Filament  voltage  to  the  12AL5 
is  reduced  to  10.5  volts  to  provide  better  noise 
limiting  action. 

Two  tubes  are  used  in  the  audio  system  of 
the  transceiver.  A 12AX7  dual  triode  serves 
as  a two  stage  resistance  coupled  speech  am- 
plifier. Only  one  section  of  the  12AX7  is 
used  for  reception  purposes.  The  dual  triode 
is  capacity  coupled  to  a 12AQ5  audio  output 
stage.  A tapped  output  transformer  matches  the 
1 2AQ5  to  the  r-f  amplifier  plate  circuit  of  the 
transmitter  section,  and  a low  impedance  wind- 
ing on  the  transformer  drives  an  external 
speaker  for  reception.  The  screen  voltage  of 
the  12AQ5  is  reduced  for  receiving  operation 
to  conserve  power  supply  drain. 

The  transmitter  section  of  the  transceiver  uses 
two  pentode  tubes.  The  oscillator  is  a 6CL6 
in  a "hot  cathode”  circuit  employing  25  Me. 
overtone  crystals.  Switch  Ss  in  the  grid  circuit 
of  the  oscillator  permits  selection  of  an  ex- 
ternal variable  frequency  oscillator.  The  plate 
circuit  of  the  oscillator  (G-Ls)  is  tuned  to  the 
50  Me.  second  harmonic  of  the  crystal.  A 
switch  (Si)  in  the  oscillator  screen  circuit 
permits  the  operator  to  turn  on  the  oscillator 
stage  during  reception  for  "zero-beat”  or  fre- 
quency marking  purposes.  The  oscillator  is  in- 
ductively coupled  to  a 5763  r-f  amplifier 


R-F  AMP.  MIXER  I'F  1*F  DET-.A-N-L 

{SOMC.)  it  MC.)  (JMC.) 


554  Receivers  and  Transceivers 


THE  RADIO 


® 

ANTENNA 

CHANGEOVER 


RYl 


TO  TRANSMITTER  OUTPUT  CIRCUIT 
TO  J 1 (ANTENNA  ) 

TO  RECEIVER  INPUT 


(D 

5763  SCREEN 
CONTROL 


TO  5763  SCREEN  CIRCUIT 
TO  RYa  -SECTION  C 


SPEAKER 

CONTROL 


Q r— ©OPEN 
Z.*  - 


TO  SPEAKER  CIRCUIT 
TO  T1  SECONDARY 


TO  GAIN  CONTROL  . Rs 
TOGRID,  iaAX7 
TO  VOLUME  CONTROL,  Ra 

TO  laAQS  SCREEN 

a+  aso 

TO  47  K RESISTOR 

B-I-TO  6CL6,  RY1  SECTION  B 
BT  350 

B-f  TO  RECEIVER 


S3A  S3B 


(.see  FiiuKE  sa ) 


Figure  31 

RELAY  AND  METER  CIRCUITS 

Relays  are  shown  in  unenergized  position. 


Stage,  which  is  plate  modulated  by  the  12AQ5 
audio  amplifier.  The  r-f  amplifier  is  bridge 
neutralized  by  capacitor  Cm  for  greatest  sta- 
bility at  the  operating  frequency  range.  The 
plate  circuit  of  the  5763  is  pi-coupled  to  a 
52-  or  72-ohm  external  load. 

Transfer  from  reception  to  transmission  is 
accomplished  by  means  of  two  d-c  relays  ac- 
tuated by  a push-to-talk  switch  in  the  micro- 
phone. The  six  volt  relay  coils  are  connected 
in  series  for  twelve  volt  operation.  Relay 
change-over  operation  is  shown  in  figure  31. 
Relay  RY-1  actuates  antenna  changeover, 
screen  control,  and  speaker  control  circuits. 
Relay  RY-2  actuates  audio  control,  12AQ5 
screen  control,  and  B-plus  control  circuits. 
Both  relays  are  shown  in  "receive”  position  in 
figures  31  and  32. 

Proper  tuning  of  the  r-f  amplifier  is  ac- 
complished by  a simple  crystal  diode  voltmeter 
(1N38A)  connected  in  the  antenna  circuit. 


The  stage  is  tuned  for  maximum  reading  of 
the  voltmeter. 

A two  pole,  four  position  rotary  switch  ( S3 ) 
serves  to  place  the  0-1  d-c  milliammeter  across 
the  appropriate  circuits  in  the  transceiver.  In 
position  1 (Ms)  the  meter  serves  as  a 0-500 
volt  meter  measuring  the  transceiver  plate 
voltage.  Position  2 (M3)  converts  the  meter 
into  a 0-10  d-c  milliammeter  measuring  power 
amplifier  grid  current.  Position  3 (M4)  moni- 
tors the  r-f  voltmeter  output,  and  position  4 
(Ml,  M2)  places  the  meter  in  the  a-v-c  con- 
trolled B-plus  line  of  the  receiver  where  it 
serves  as  a signal  strength  meter. 

Under  normal  driving  and  charging  condi- 
tions, the  voltage  of  the  standard  ”12  volt” 
battery  can  vary  from  11  volts  to  14.5  volts. 
This  fluctuation  can  raise  havoc  with  the  re- 
ceiver oscillator  stability.  It  is  annoying,  to 
say  the  least,  to  have  the  receiver  calibration 
change  with  variations  of  car  speed  and  gen- 
erator voltage.  To  eliminate  this  effect,  an 
Amperite  ballast  tube  (R4)  is  placed  in  series 
with  the  filament  circuit  of  the  6CB6  receiver 
oscillator  tube.  This  current  sensitive  device 
compensates  for  voltage  changes  in  the  fila- 
ment circuit,  holding  the  voltage  at  the  fila- 
ment terminals  of  the  oscillator  tube  within 
a few  percent  of  a nominal  6.3  volts. 

The  six  volt  tubes  of  the  receiver  are  con- 
nected in  series  parallel  for  operation  across 
the  12-volt  power  supply.  A compensating  re- 
sistor is  required  across  the  filament  of  the 
5763  to  equalize  the  current  drain  with  that 
of  the  6CL6  oscillator. 

Transceiver  Layout  Figures  29,  33,  and  34 
and  Assembly  provide  a general  lay- 

out of  the  transceiver. 
The  unit  is  built  upon  an  aluminum  chassis 
measuring  9"  x 7"  x 1^/i"  in  size.  This  assem- 
bly fits  within  a steel  wrap-around  type  cab- 
inet 4Vi''  high,  and  approximately  the  same 
length  and  depth.  The  cabinet  is  formed  of 
perforated  metal  to  ensure  proper  ventilation 
■^hile  at  the  same  time  reducing  spurious  radi- 
ation to  a minimum. 

Viewed  from  the  top  rear  ( figure  33),  the 
receiver  portion  of  the  unit  occupies  the  right 
half  of  the  chassis  and  the  transmitter  and 
modulator  occupy  the  left  half.  The  external 
plugs  and  receptacles  ate  mounted  on  the 
rear  apron  of  the  chassis  as  is  the  modulation 
gain  control,  R3.  These  components  project 
through  a cut-out  in  the  rear  of  the  steel  en- 
closure. All  the  ceramic  variable  capacitors  that 
serve  as  circuit  adjustment  trimmers  are  flush 


Li  6CB6R-F  L2  dCBdmix. 

(S0~S4MC.)  JS0-S4MC.)  


T 1 33 

(fMC.)  "S' 


12AQ&  250 


¥;  - S££  FieufiE  35 


*^13  10JJFD 

X. 


C12A^  10J<  ^Cl2B 
lOJUFD^  1.0  lOjJFD 
450  " " 450 


SOLD£R  S/X  6.3  VOLT,  150  MA. 
BULBS  TOGETHER.  JUMPER 
CENTER  TERMINALS  TOGETHER. 


ISg  6CL6< 
2WT 


Figure  32  ~ 

SCHEMATIC,  SIX  METER  TRANSCEIVER 

Coii  data  and  i-f  transformer  modification  is  given  in  figure  35.  Relay  and  meter  connections  are  given  in  figure  31. 

Ci~Cs-C5-—6-30  fififd.  Centralab  827C  RYi,  RYi — 3 pole,  double  throw  with  6.3  volt  d~c  coil.  Advance  #MG-3C6VA 

C4-C7-SS-C9— 3-12  /i/x/d.  Centralab  827B  Ti-Ts — 1500  fcc.  i-f  transformer,  J.  W.  Miller  Co.  12-W1 

Cs-Ci — 15  fi/ifd,  per  section.  All-Star  Products  Co.,  Delionce,  Ohio.  Model  C3  (see  figure  35  for  modification) 

Citf—l  fififd.  variable  "piston-type'\  Erie  5326  T« — 5 K pri.,  6.7  K $ec.,  and  4 ohm  winding.  Triad  M5-Z 

^ ..„tA  Ufintmnriund  HF-3S  Ri—^Amperiie  Ballast  Tube  #3Tf-4A 

— * — — » F'flUnwwtin  Chassis  Co.  #LTC-464 


556  Receivers  and  Transceivers 


THE  RADIO 


mounted  to  the  chassis  and  may  be  tuned 
from  the  top  of  the  unit,  with  the  exception 
of  the  receiver  oscillator  trimmer  Ci  which  is 
mounted  below  chassis  to  conserve  space. 

Below  the  chassis,  a small  aluminum  shield 
measuring  I/2”  long  and  I/4"  high  is  positioned 
between  the  oscillator  output  coil  Lj  and  the 
receiver  r-f  amplifier  plate  coil  !•.  This  is  the 
only  interstage  shield  required  in  the  trans- 
ceiver. The  two  transfer  relays  RY-1  and  RY-2 
are  mounted  in  the  under-chassis  area  be- 
neath the  modulation  transformer  Tg.  Between 
the  relays  is  placed  an  8-terminal  phenolic 
mounting  strip.  To  the  side  of  this  strip  is 
placed  the  midget  S-meter  potentiometer  Ri. 
All  sockets  and  trimming  capacitors  are  mount- 
ed in  place  using  4-40  hardware  with  solder- 
ing lugs  placed  beneath  the  nuts  in  various 
convenient  locations. 

Transceiver  The  wiring  of  the  unit  is  quite 
Wiring  simple  if  done  in  the  proper 

sequence.  The  underchassis  area 
contains  many  small  components  but  these  need 
not  be  crowded,  provided  proper  care  is  taken 
in  the  layout  and  installation  of  parts.  Be  sure 
the  i-f  transformers  have  been  modified  accord- 


ing to  the  data  of  figure  35  before  wiring  is 
started. 

The  majority  of  components  are  installed 
between  the  socket  pins  of  the  various  tubes 
and  five  large  phenolic  terminal  strips.  One  of 
these  strips  (previously  mentioned)  is  placed 
between  the  two  transfer  relays.  A four  term- 
inal and  a six  terminal  tie-point  strip  are 
mounted  in  line  running  parallel  to  the  i-f 
tube  sockets  of  the  receiver,  towards  the  chas- 
sis-edge side  of  the  sockets.  The  i-f  stage  com- 
ponents mount  between  these  tie-points  and 
the  pins  of  the  6BJ6  and  12AL5  tube  sockets. 
A fourth  terminal  strip  (8  tie-points)  is  par- 
allel to  and  directly  beside  the  edge  of  the 
chassis  adjacent  to  the  r-f  tube  sockets  of  the 
receiver.  The  smaller  components  of  the  r-f 
section  mount  between  this  strip  and  the  pins 
of  the  6CB6  sockets. 

The  r-f  coils  of  the  receiver  section  (Li,  L2, 
Ls,  and  L4)  are  mounted  directly  to  the  ceramic 
trimming  capacitors  associated  with  each  coil. 
The  interstage  coupling  coils  (Ls  and  Le)  of 
the  transmitter  section  are  mounted  to  their 
respective  trimming  capacitors  and  may  be 
seen  in  the  center  of  the  chassis  in  figure  34. 


Figure  33 
REAR  VIEW  OF 
SIX  METER 
TRANSCEIVER 
CHASSIS 

Receiver  section  of  unit 
is  at  right  with  irimmef 
capacitors  seen  on  chas~ 
sis  deck.  Center  section 
of  three  gang  variable 
capacitor  is  unused. 
Transmitter  output  con- 
trols  are  located  at  up- 
per left.  Filament  volt- 
age regulator  tube  is 
at  rear  of  chassis,  just 
above  modulation  level 
control  potentiometer. 
5763  amplifier  tube  is 
behind  modulation  trans- 
former, with  shielded 
6CL6  tube  to  right.  I-f 
strip  runs  along  center 
of  chassis.  Antenna  and 
speaker  plugs  and  power 
receptacle  are  along  rear 
lip  of  chassis.  Receiver 
audio  control  is  at  upper 
right  of  front  panel. 


HANDBOOK 


6-Meter  Transceiver  557 


After  all  major  components  have  been 
mounted  to  the  chassis  the  various  socket 
ground  connections  should  be  made,  followed 
by  the  filament  wiring.  The  speech  amplifier 
should  be  wired  next,  taking  care  to  place 
small  components  directly  between  socket 
pins  to  conserve  space.  All  the  small  compon- 
ents of  the  i-f  system  can  be  mounted  above 
the  tube  sockets,  and  in  the  space  between  the 
sockets  and  the  tie-point  strips.  The  power 
leads  from  the  wired  sections  of  the  transceiver 
should  now  be  run  to  the  two  relays. 

The  next  step  is  to  complete  the  receiver 
wiring.  Shielded  leads  are  run  to  the  volume 
control  Rj  mounted  in  the  upper-left  hand 
corner  of  the  receiver  panel  (as  viewed  from 
the  front).  The  S-meter  wiring  circuit  should 
be  completed  as  the  final  wiring  step.  Note 
that  the  case  of  potentiometer  Ri  is  "hot”  and 
should  be  insulated  from  the  chassis  with  fibre 
bushings. 

Receiver  The  receiver  may  be  tested  be- 

Check-out  fore  the  transmitter  wiring  is 

done.  Relays  RY-1  and  RY-2 
are  in  an  unenergized  position  during  recep- 
tion. Examine  both  relays  to  make  sure  that 
the  back  contacts  (normally  closed)  are  in 
good  operating  condition. 


FIGURE  35 

COIL  TABLE  FOR  SIX  METER  TRANSCEIVER 
Li,  L2,  L4  - 8 turns  N®18  E..  3/8"  DIAM.,  1/2"  LONG  . 

L3-7  TURNS  N<*16  E.,  1/2"  DIAM.,  1/2"  LONG- 
(Btty  COIL  STOCK  ) 

Ls,  L6  - 10  TURNS  N®  22  E.,  5/16"  DIAM.,  5/16"  LONG. 

ADJUST  SPACING  BETWEEN  COILS  FOR  MAXI- 
MUM GRID  DRIVE.  COILS  MOUNT  END  TO  END. 
L7-6  TURNS  N®  16  E.,  1 /2  " Dl  AM ..  5/8  " LON  G.  AIRWOUND. 

Ti.TS-REPLACE  IOOUUFD  padding  CAPACITORS 

WITH  200UiJFD  CAPACITORS  TO  RESONATE 
Wl  NDINGS  TO  1 MC. 


A pov/er  supply  capable  of  delivering  250 
volts  at  100  milliamperes  and  12.6  volts  at 
3 amperes  is  required  to  test  the  transceiver. 
A 4-ohm  loudspeaker  is  also  required.  The 
receiver  is  energized  and  the  important  voltage 
points  are  checked  with  a high  resistance  volt- 
meter and  should  comply  closely  with  the 
values  of  figure  32.  When  the  receiver  section 
is  determined  to  be  in  operational  order,  a 
tone  modulated  1 megacycle  signal  should 
be  loosely  coupled  to  the  plate  circuit  of  the 
6CB6  mixer  tube  (pin  5)  and  the  slug  cores 
of  all  i-f  transformers  tuned  for  maximum 
audio  signal.  As  the  stages  are  bought  into 
resonance,  the  input  signal  to  the  mixer  tube 
should  be  reduced  to  prevent  over-excitation 
of  the  stages. 


Figure  34 
UNDER-CHASSIS 
VIEW  OF 
TRANSCEIVER 

Changeover  relays  are 
located  at  the  left  edge 
of  the  chassis.  The  hf 
amplifier  runs  down  the 
center  of  the  chassis, 
with  the  receiver  r-f 
stages  to  the  right. 
Transmitter  stages  are 
located  between  front  re- 
lay and  i-f  strip.  Audio 
stages  are  placed  along 
rear  of  chassis.  Small 
components  are  mounted 
between  socket  pins  and 
adjacent  phenolic  termi- 
nal strips. 

Note:  filament  circuit  is 
designed  for  d.c.  opera- 
tion. If  a.c.  filament  op- 
eration is  desired,  high 
resistance  relay  coils  op- 
erated from  the  hiah 
voltage  supply  should  be 
employedf 


558  Receivers  and  Transceivers 


THE  RADIO 


The  next  step  is  to  check  the  tuning  range 
of  the  high  frequency  oscillator,  which  should 
be  17.0 — 17.7  Me.  The  range  may  be  checked 
with  an  accurately  calibrated  receiver  or  fre- 
quency meter.  The  fixed  trimming  capacitor 
determines  the  overall  frequency  span  of  the 
oscillator,  and  padding  capacitor  Ci  determines 
the  range  setting.  Padding  capacitor  Q is  ad- 
justed so  that  17-0  Me.  occurs  when  the  main 
tuning  dial  is  set  to  five  degrees,  allowing  a 
slight  amount  of  "overlap”  at  the  band  edges. 
As  the  main  tuning  capacitor  setting  is  de- 
creased the  oscillator  will  tune  to  17.7  Me. 
If  it  is  desired  to  widen  or  diminish  the  tuning 
range,  the  value  of  the  series  capacitor  may  be 
changed  slightly.  Once  the  proper  oscillator 
range  has  been  established,  the  tripler  circuit 
should  be  adjusted  to  track  with  it.  The  main 
tuning  capacitor  should  be  set  to  17.7  Me.  and 
padding  capacitor  G adjusted  so  that  the  tuned 
circuit  resonates  at  53  Me.,  as  checked  with  a 
grid  dip  oscillator.  The  main  tuning  dial  is 
then  reset  to  place  the  oscillator  on  17-0  Me.; 
trimming  capacitor  G is  adjusted  so  that  the 
tripler  circuit  resonates  at  51  Me.  Once  the 
two  circuits  have  been  brought  into  rough 
alignment  they  may  be  placed  "on  the  nose" 
by  adjusting  G and  G for  constant  grid  cur- 
rent in  the  mixer  stage  across  the  tuning 
range.  Temporarily  unsolder  the  grounded  ter- 
minal of  the  0.1 -megohm  grid  resistor  of  the 
6CB6  mixer  tube  (pin  1)  and  place  a 0-100 
micro-ammeter  in  series  with  the  resistor  to 
ground.  Adjust  Ci  at  the  high  frequency  end 
of  the  tuning  dial  and  G at  the  low  frequency 
end  until  the  rectified  grid  current  is  relatively 
constant  across  the  band.  Under  proper  oper- 
ating conditions,  the  reading  should  be  about 
15  microamperes. 

The  last  step  is  to  align  the  tuned  circuits 
of  the  r-f  stage.  These  are  relatively  broad  band 
and  need  only  be  peaked  over  the  usual  fre- 
quency range  of  operation.  For  most  purposes 
they  should  be  resonated  at  50.5  Me.  for  opti- 
mum operation  at  the  low  frequency  end  of 
the  band.  Signals  above  52  Me.  will  be  slightly 
attenuated  by  this  adjustment.  If  operation  at 
the  high  frequency  end  of  the  six  meter  band 
is  desired,  the  circuits  should  be  peaked  at  53 
Me.  with  a consequent  slight  reduction  of  per- 
formance in  the  50  Me.  region. 

Completion  of  the  Once  the  receiver  is 

Transmitter  Section  judged  to  be  in  good 

operating  condition,  at- 
tention should  be  turned  to  the  transmitter. 
The  lead  from  the  XTAL-VFO  switch  Sj  to 
the  v-f-o  input  receptacle  J=  on  the  rear  of  the 


chassis  should  be  made  of  coaxial  cable,  as 
should  the  lead  from  relay  RY-1  to  the  an- 
tenna receptacle  Ji.  In  addition,  the  lead  from 
the  receiver  antenna  input  circuit  to  relay  RY-1 
should  be  shielded. 

The  majority  of  small  components  of  both 
the  oscillator  and  amplifier  stages  may  be 
mounted  by  their  leads  between  socket  pins 
and  a nearby  5-terminal  phenolic  tie-point 
strip.  The  plate  coil  L-  of  the  5763  amplifier 
stage  is  mounted  above  chassis,  directly  be- 
hind amplifier  tuning  capacitors  Cn  and  Cu. 
Small  National  Co.  polystyrene  feed  through 
bushings  are  used  to  support  the  coil.  Neutral- 
izing capacitor  Go  is  soldered  to  a rotor  termi- 
nal of  plate  tuning  capacitor  Cn.  The  lead 
from  Cio  to  the  "cold”  end  of  Lo  passes  through 
a /Tinch  hole  drilled  in  the  chassis.  To  insure 
a good  ground  return,  the  rotors  of  Cn  and 
Cn  are  connected  to  the  chassis  with  a short, 
heavy  copper  strap. 

The  final  step  is  to  complete  the  wiring  to 
the  meter  switch  and  the  relays  as  outlined  in 
figure  31.  When  this  is  completed,  all  the 
transceiver  wiring  should  be  completely  check- 
ed and  the  chassis  thoroughly  cleaned  of  loose 
bits  of  solder,  wire,  etc. 

Transmitter  It  is  necessary  to  close  both  re- 
Adjustment  lays  to  test  the  transmitter  op- 
eration. For  the  time  being, 
they  may  be  wedged  close  with  a bit  of  match- 
stick  if  a 12-volt  d-c  supply  is  not  readily 
available.  A 25  Me.  crystal  should  be  plugged 
in  the  panel  holder  and  switch  S2  set  to  the 
XTAL  position.  Switch  Si  is  set  to  OSC  posi- 
tion, and  the  lead  from  the  screen  circuit  of 
the  5763  amplifier  tube  to  the  contact  of  re- 
lay RYl-B  is  temporarily  opened,  removing 
screen  voltage  from  the  r-f  amplifier.  Filament 
and  plate  voltage  is  applied  to  the  transceiver 
and  capacitor  G is  tuned  for  oscillation  of  the 
crystal.  The  meter  of  the  transceiver  is  switched 
to  the  GRID  (Ms)  position  and  capacitor  G 
adjusted  for  maximum  grid  current.  A reading 
of  half-scale  should  be  obtained.  The  coupling 
between  Ls  and  Ls  can  be  varied  slightly  to 
obtain  the  proper  reading. 

The  dummy  antenna  (figure  32)  is  next 
plugged  into  antenna  receptacle  Ji  and  loading 
capacitor  Gs  set  at  full  capacity.  The  plate  cir- 
cuit of  the  amplifier  stage  should  be  resonated 
at  50  Me.  with  the  aid  of  a grid  dip  oscillator 
coupled  to  coil  Lt.  The  turns  of  L?  should  be 
adjusted  so  resonance  occurs  at  half-capacity 
setting  of  tuning  capacitor  Cn.  After  this  pre- 
liminary adjustment  has  been  made,  capacitor 
Cn  should  be  tuned  through  its  range  while 


HANDBOOK 


28-Mc.  Transceiver  559 


carefully  noting  the  grid  current  of  the  ampli- 
fier. Unless  neutralizing  capacitor  Go  has  been 
correctly  set  by  a lucky  accident,  the  grid  cur- 
rent reading  will  show  an  abrupt  kick  as  ca- 
pacitor Cii  is  tuned  through  resonance.  The 
setting  of  the  neutralizing  capacitor  should  be 
slowly  varied  so  as  to  minimize  the  kick  of 
grid  current.  After  change  of  setting  of  Go, 
grid  tuning  capacitor  Gi  should  be  retuned  for 
maximum  grid  current.  An  adjustment  of  Go 
will  soon  be  found  that  will  reduce  the  grid 
current  variation  to  an  imperceptible  kick  of 
the  meter.  When  this  point  is  found,  the  rotor 
of  Go  should  be  fastened  in  place  with  a drop 
of  nail  polish  and  the  screen  lead  to  RY-1 
should  be  resoldered. 

Switch  Si  is  now  set  to  the  transmit  posi- 
tion and  power  is  again  applied  to  the  trans- 
mitter. The  meter  switch  is  placed  in  the  OUT- 
PUT (Ml)  position,  and  plate  tuning  capaci- 
tor Gi  adjusted  for  maximum  meter  reading. 
At  the  same  time,  the  bulbs  of  the  dummy  an- 
tenna should  glow.  Capacitors  Cn  and  Cu  are 
touched  up  for  maximum  indication  of  the 
meter. 

In  order  to  test  the  modulator  section  of  the 
transmitter  when  using  an  a-c  filament  supply, 
it  is  necessary  to  disconnect  the  push-to-talk 
circuit  at  pin  #2  of  the  microphone  plug; 
otherwise  there  will  be  a loud  hum  on  the 
audio  signal.  When  this  is  done,  the  micro- 
phone may  be  plugged  into  the  jack  and  the 
gain  control  (Ri)  advanced.  The  dummy  an- 
tenna should  increase  in  brilliance  under  pro- 
per modulation.  If  the  coupling  to  the  dummy 
load  is  too  tight  downward  modulation  will 
result,  and  the  bulb  brilliance  will  drop.  The 
capacity  of  loading  capacitor  Cm  should  be  in- 
creased, dropping  the  degree  of  loading. 

The  transmitter  is  designed  to  operate  into 
a coaxial  transmission  line  having  an  imped- 
ance value  between  50  and  75  ohms.  When  a 
high  value  of  standing  wave  ratio  exists  on 
the  line  the  impedance  presented  to  the  pi- 
network  amplifier  may  be  of  such  magnitude 
as  to  preclude  the  possibility  of  a proper 
match.  It  is  permissible  to  add  a second  mica 
capacitor  in  parallel  with  loading  capacitor  Ct2 
to  extend  the  operating  range  of  the  pi-net- 
work. If  more  than  50  g/afd.  has  to  be  added, 
the  SWR  ratio  should  either  be  lowered  by 
proper  adjustments  to  the  antenna,  or  the 
length  of  the  transmission  line  between  the 
transciver  and  antenna  should  be  altered  in 
order  to  present  a more  reasonable  load  to 
the  amplifier  circuit.  Maximum  carrier  power 
output  when  properly  tuned  is  about  41^  watts. 


All  r-f  adjustments  should  be  carried  out  with 
the  purpose  of  obtaining  maximum  reading  on 
the  output  meter  of  the  transmitter  under  a 
given  set  of  antenna  conditions.  Under  these 
conditions,  maximum  transceiver  plate  current 
is  100  milliamperes  at  a plate  potential  of 
250  volts. 

The  Meter  A final  word  should  be  said 
Circuit  about  the  metering  circuit  of  the 

transceiver.  The  particular  meter 
used  has  a d-c  resistance  of  100  ohms,  and  a 
full  scale  reading  of  one  milliampere.  In  order 
to  simplify  the  problem  of  obtaining  meter 
shunts,  the  meter  is  converted  into  a 0-1  volt- 
meter by  adding  a 900  ohm  resistor  in  series 
with  the  meter  in  the  grid  current  position 
(Ms).  The  meter  is  placed  across  a 1000  ohm 
shunt  in  this  position.  Since  the  resistance  of 
the  meter  circuit  and  the  shunt  ate  equal,  one- 
half  the  grid  current  flows  through  the  meter, 
and  one-half  through  the  shunt.  Thus,  when 
the  meter  reads  full  scale  (one  milliampere), 
two  milliamperes  of  grid  current  are  flowing 
in  the  grid  return  circuit. 

When  the  meter  is  switched  to  S-METER 
position  (Ml,  Ms)  of  switch  Ss,  the  meter  is 
placed  in  a bridge  circuit,  the  variable  leg  of 
which  is  the  internal  resistance  of  the  a-v-c 
controlled,  i-f  amplifier  tube.  The  bridge  is 
balanced  by  proper  setting  of  Ri;  meter  sensi- 
tivity is  set  by  the  value  of  the  series  resistor 
(2.2K).  Dropping  the  value  of  this  resistance 
will  increase  the  sensitivity  of  the  meter. 

When  S.i  is  set  to  the  VOLTS  position,  the 
meter  is  placed  in  series  with  a 500  K resistor 
(Ms)  and  is  converted  into  a 0-500  volt  meter 
to  check  the  operating  potential  of  the  plate 
supply. 

27-7  A "Hot" 

Transceiver 
for  28  Megacycles 

This  little  transceiver  is  vivid  proof  that 
DX  can  be  easily  worked  on  the  10  meter 
band.  During  a test  period  of  two  weeks,  using 
a dipole  antenna,  over  80  foreign  phone  con- 
tacts were  made  in  four  continents,  including 
three  contacts  with  India.  Power  input  to  the 
transmitter  section  of  the  unit  was  under  10 
watts.  Admittedly,  such  DX  is  not  a daily 
feature  with  such  low  power,  but  it  does 
prove  that  a few  watts  into  a well  situated  an- 
tenna can  work  wonders  on  the  28  megacycle 
band. 

The  28-10  transceiver  (28  megacycles,  10 
watts)  is  designed  to  operate  over  the  range 


of  28.0-29.7  Me.  The  receiver  portion  em- 
ploys five  tubes  in  a single  conversion  cir- 
cuit. Tvifo  tubes  do  double  duty  as  the  audio 
section  for  reception  and  the  modulator  for 
transmission.  Four  tubes  are  used  in  the  r-f 
section,  making  a total  of  eleven  tubes  plus  a 
voltage  regulator.  A voltage  doubler  selenium 
rectifier  supplies  plate  voltage  for  the  complete 
transceiver.  The  whole  unit  weighs  less  than 
eleven  pounds  so  it  is  readily  transportable, 
even  by  air.  The  power  source  is  115  volts, 
50-60  cycles,  a-c. 


Figure  36 

TEN  METER  TRANSCEIVER  HAS 

SIMPLE  CONTROL  PANEL 

Used  with  a dipole  antenna,  this  10  watt 
28  Me.  transceiver  has  made  three  contacts 
with  India.  Main  tuning  dial  of  receiver  is 
at  center  with  send-receive  switch  below  it. 
At  upper  right  is  combination  S-meter  and 
transmitter  tuning  indicator.  Transmitter 
tuning  controls  are  at  left  with  "push-to- 
xero"  switch  near  receiver  tuning  dial. 
Unit  features  self-contained  power  supply 
for  ns  volt  operation. 


Transceiver  A block  diagram  of  the  trans- 
Circuit  ceiver  is  shown  in  figure  37. 

Change-over  from  receive  to 
transmit  is  accomplished  by  a ceramic  rotary 
switch.  Si.  The  receiver  employs  one  r-f  stage 
and  two  i-f  stages  for  optimum  selectivity  and 
sensitivity.  The  r-f  stage  employs  a low  drain 
6BJ6  remote  cut-off  pentode  with  tuned  grid 
and  tuned  plate  circuits.  The  plate  of  the  r-f 
tube  is  shunt  fed  through  a 10  K composition 
resistor  and  the  tuned  circuit  is  placed  in  the 
low  potential  grid  circuit  of  the  next  stage.  A 
6U8  is  employed  here  as  a combined  mixer 
and  oscillator.  Grid  circuit  injection  is  used. 
The  triode  section  of  the  6U8  serves  as  the 
local  oscillator,  tuning  from  26.5-28.2  Me.  for 
10  meter  coverage.  A three  section  variable 
capacitor  ClA-B-C  simultaneously  tunes  the 


Figure  37 

BLOCK  DIAGRAM  OF  TEN  METER  TRANSCEIVER 


HANDBOOK 


28-Mc.  Transceiver  561 


r-f,  mixer,  and  oscillator  stages.  Series  padding 
in  the  r-f  and  mixer  circuits  is  employed  for 
good  tracking  across  the  band. 

An  i-f  channel  of  1.5  Me.  is  used  in  this 
unit.  This  particular  frequency  was  chosen  to 
eliminate  the  troublesome  "image”  problem, 
so  prevalent  on  the  10  meter  band  when  the 
usual  455  kc.  i-f  channel  is  employed.  Selec- 
tivity suffers  a hit  when  the  higher  frequency 
channel  is  used,  but  a total  of  eight  tuned 
circuits  (four  transformers)  provide  an  ac- 
ceptable passband  for  voice  reception.  A 6AL5 
double  diode  serves  as  the  detector,  a-v-c  rec- 
tifier, and  automatic  noise  limiter.  The  a-n-1 
is  left  in  the  circuit  at  all  times. 

The  audio  section  of  the  28-10  uses  two 
tubes.  A 6BJ6  serves  as  a resistance  coupled 
pentode  voltage  amplifier  which  is  coupled  to 
a 6AQ5  power  amplifier.  A tapped  output 
transformer  matches  the  6AQ5  to  the  r-f  am- 
plifier plate  circuit  of  the  transmitter  section, 
and  a low  impedance  winding  on  the  trans- 
former matches  the  audio  system  to  external 
earphones  or  speaker.  The  input  circuit  of  the 
6BJ6  forms  an  R-C  dividing  network  isolating 
the  microphone  input  from  the  audio  output 
circuit  of  the  receiver.  No  switching  is  required 
in  this  circuit. 

The  transceiver  is  designed  to  be  used  with 
a low  impedance  carbon  microphone.  Voltage 
for  microphone  operation  is  taken  from  a tap 
on  the  cathode  circuit  of  the  6AQ5  amplifier 
stage.  This  circuit  is  broken  by  switch  SID 
during  reception.  At  the  same  time,  switch  sec- 
tion SIE  connects  the  earphone  circuit  to  the 
low  impedance  winding  on  the  output  trans- 
former Te. 

Three  tubes  plus  a gas-type  voltage  regula- 
tor and  a diode  t-f  indicator  ate  used  in  the 
transmitter  portion  of  the  transceiver.  A 6BJ6 
serves  as  a "Tri-tet”  oscillator  coveting  the 
range  of  7.0-7.25  Me.  The  plate  circuit  is  slug- 
tuned  to  the  14  Me.  second  harmonic.  A 
switch  (S=)  in  the  plate  circuit  of  the  oscillator 
permits  the  operator  to  turn  on  this  stage  for 
"zero-beat”  or  frequency  marking  purposes- 
The  oscillator  is  capacitively  coupled  to  a 
6AQ5  doubler  stage  whose  plate  circuit  is 
tuned  to  the  28  Me.  region.  Switch  segment 
SIC  removes  the  plate  voltage  from  this  stage 
during  reception  and  applies  it  to  the  r-f  sec- 
tion of  the  receiver.  The  plate  circuit  of  the 
6AQ5  doubler  stage  is  capacitively  coupled 
to  a second  6AQ5  serving  as  the  modulated 
amplifier.  This  latter  stage  is  bridge  neutralized 
by  capacitor  Cs  for  stability  at  the  operating 
frequency.  The  plate  circuit  of  the  amplifier 


is  a simple  pi-network  designed  to  match  loads 
of  50  to  75  ohms.  The  external  antenna  is 
switched  between  the  transmitter  and  the  re- 
ceiver by  means  of  switch  segment  SIA.  A 
separate  section  of  the  transfer  switch  (SIB) 
removes  plate  voltage  from  the  6AQ5  r.f. 
amplifier  stage  and  applies  an  audio  load  to  the 
modulator  during  reception.  This  auxiliary 
loading  prevents  audio  feedback  when  the 
headphones  are  removed  from  the  jack  Jj,  and 
permits  the  use  of  headphones  of  any  imped- 
ance to  be  used  without  the  danger  of  spurious 
feedback.  The  overall  gain  of  the  audio  sys- 
tem is  quite  high  and  it  is  built  in  a small 
space.  As  a result,  it  does  not  have  the  reserve 
of  stability  that  would  normally  be  expected. 
The  switching  system,  too,  tends  to  create 
small  feedback  loops  that  must  be  carefully 
controlled.  The  chief  cause  of  audio  instability 
(or  feedback)  in  this  case  is  due  to  the  close 
proximity  of  transformers  Ta  and  Ta  above  the 
chassis.  Feedback  can  be  reduced  or  enhanced 
by  reversing  the  polarity  of  the  secondary 
winding  of  Ta.  The  additional  audio  filtering 
shown  in  the  schematic  of  figure  38  reduced 
this  tendency  to  a minimum.  As  a final  pre- 
caution, a small  shield  (cut  from  a segment  of 
tin  can)  was  soldered  to  the  top  of  the  core 
of  Ta.  The  shield  projected  downwards  over 
the  windings  of  the  transformer,  as  seen  in 
Figure  39-  A more  expensive  solution  would 
have  been  to  employ  a shielded  transformer  in 
this  portion  of  the  circuit.  A similar  shield  is 
soldered  to  the  core  of  the  filament  transform- 
er to  reduce  hum  pickup  by  the  adjacent  audio 
transformers. 

The  a-c  operated  power  supply  employs  two 
"replacement  type”  selenium  rectifiers  in  a 
voltage  doubling  circuit  delivering  250  volts 
at  100  milliamperes.  A small  filament  trans- 
former (Tt)  provides  6.3  volts  for  the  tubes 
and  pilot  lamp.  This  particular  voltage  doubler 
has  the  negative  side  of  the  high  voltage  in 
common  with  one  side  of  the  primary  line. 
As  a result,  there  is  a fifty-fifty  chance  that 
the  chassis  of  the  transceiver  will  be  "above 
ground”  by  the  amount  of  the  line  voltage, 
which  in  this  case  is  115  volts.  This  can  be  a 
lethal  situation  if  permitted  to  exist.  A prac- 
tical solution  is  to  remove  the  "ground”  side 
of  the  transmitter  power  cable  from  the  usual 
two  wire  line,  and  connect  it  directly  to  the 
external  ground,  as  shown  in  figure  4l.  This 
ensures  that  the  chassis  of  the  transceiver  is  at 
ground  potential.  The  "hot”  side  of  the  line 
returns  to  one  pin  of  the  line  plug.  If  the  plug 


ooooonn 


Figure  38A 

SCHEMATIC,  TO  METER  TRANSCEIVER 

— .01  fifd.,  600  vo/t  ceramic  capacitor.  Ceniralab  006-/03  RFC — 2^  mh.  Millen  miniaturized  J300-2500 

lA-B-C — 75  ji^fd.  per  section.  All  Star  Products  Co.,  Defiance,  Ohio,  Type  C3  SM-B-C-0-£-f— 6 pole,  two  position  switch.  Centralab  PA~2019 

t — 15  fxiiid.  Bud  MC-565  SRi,  5Rt—^150  ma.  selenium  rectifier.  Federal  1005 

s — 30  /x/ifd.  Bud  LC-765/  TfTi — 7500  kc.  i-f  transformer.  J.  W.  Miller  Co.  12-W1 

,, — 400  fi/xfd.  Allied  Radio  Co.,  Chicago,  III.  #67H-009  Ts — 700  ohms  pri.  to  grid.  Triad  A~5X 

-7  nixfd.  Erie  532B  Ts — 5 K pri.,  6.7  K and  4 ohm  sec.  Triad  M~4Z 


Hi — 3 henry  at  100  ma.  Stancor  C-2304 


Tr— 6.3  volt,  3 amp.  Stancor  P-6466 
Chassis  and  cabinet — California  Chassis  Co.  ifLTC-470 


> 

a 


562  Receivers  and  Transceivers 


HANDBOOK 


28-Mc.  Transceiver  563 


FIGURE  38B 

COiL  TABLE,  10  METER  TRANSCEIVER 


Li  - ANTENNA:  Z TURNS  N*  16  INSULATED  WIRE 
AROUND  "COLD"  END  OF  Ll.  GRID:  14TURN3 
N*ieE.,  1/4"  DIAM.,  1/Z«LONG.  J.W.  MILLER  J447-F 
LZ-  13  TURNS  16  E.,  SAME  AS  Li 
L3-7  TURNS  N*  16  E.,  SAME  AS  Li,  TAP  AT  3 TURNS 
FROM  GROUND  END. 

L4-  13TURNS  N*  16  E.,  3/4"  DIAMETER,  1/2"  LONG. 
NATIONAL  XR~tZ  FORM. 

LS-  J.W.  MILLER  FORM  1447-F,  1/4*  OIAM.,  1/2" 
LONG  CLOSEWOUND  WITH  N*  26  E.  FULL  LENGTH 
OF  SPACE. 

L6-  14  TURNS  N*  18  E.,  SAME  AS  Li 

L7-  7 TURNS  N*  18  E.,  5/6'  DIAM.,  3/6' LONG  {B4W STOCK'S 


VOLTAGE  MEASUREMENTS 

TUBE 

PLATE 

SCREEN 

CATHODE 

6BJ6  R-F 

160 

150 

2.5 

All 

260 

185 

7 

OSC. 

100 

— 

6BJ6  l-F'S 

165 

140 

3 

6AQ5 

AUDIO 

260 

260 

TX-12.5 

RX-  17 

is  inserted  in  a wall  socket  incorrectly  nothing 
will  happen.  The  transceiver  simply  refuses  to 
function,  and  no  dangerous  situation  is  created. 
Simply  reversing  the  plug  will  energize  the 
circuit,  with  no  danger  to  the  operator.  The 
dual  power  lead  connection,  therefore,  is  a 


simple  and  foolproof  system  of  protection  and 
permits  the  use  of  the  inexpensive  and  light 
weight  selenium  supply.  The  only  other  prac- 
tical alternative  is  to  employ  a power  trans- 
former which  will  add  excessive  weight  and 
bulk  to  the  package. 

Because  this  hazard  exists  with  any  radio 
equipment  of  the  so-called  ac-dc  type,  many 
of  the  newer  buildings  are  being  wired  with 
polarized  power  receptacles  having  a separate 
grounding  pin.  The  use  of  this  new  wiring 
technique  will  help  to  reduce  the  inherent 
"shock  potential”  of  the  practical  and  efficient 
half-wave  rectifier  systems,  of  which  this  is  an 
example. 

Transceiver  Layout  The  component  place- 
and  Assembly  ment  may  be  seen  in 

figures  39  and  40.  The 
28-10  transceiver  is  built  upon  an  aluminum 
chassis  measuring  11"  x 8 Vi"  x 2"  in  size, 
and  fits  within  a steel  wrap-around  type  cabi- 
net 5%"  high.  The  cabinet  is  formed  of  per- 
forated metal  to  ensure  proper  ventilation.  Ra- 
diation of  harmonics  from  the  assembly  is  re- 
duced to  a minimum  with  this  type  of  en- 
closure. 

Layout  of  the  major  components  above  the 
chassis  may  be  seen  in  figure  39-  The  receiver 


Figure  39 
REAR  VIEW  OF 
TRANSCEIVER 
CHASSIS 

Tuning  capacitor  of  var/> 
abfe  frequency  oscitlator 
is  in  foreground  of  photo, 
panel  driven  by  an  ex- 
tension shaft.  To  the  left 
of  the  capacitor  is  the 
adjustable  slug  core  of 
the  oscillator  coll.  Along 
rear  edge  of  the  chassis 
are  the  selenium  rectifi- 
ers, the  filament  trans- 
former and  the  high  volt- 
age filter  capacitor. 
Transmitter  tubes  are  at 
right  of  chassis,  with  am- 
plifier tube  next  to 
panel.  At  far  right  Is 
voltage  regulator  tube. 
Audio  section  is  at  up- 
per left  with  9006  diode 
rectifier  in  front  of  the 
input  audio  transformer, 
t-f  amplifier  is  at  center 
chassis,  parallel  to  front 

panel. 


564  Receivers  and  Transceivers 


THE  RADIO 


section  of  the  unit  occupies  the  center  section 
of  the  chassis  with  the  power  supply  directly 
behind  it.  To  the  right  is  the  transmitter  sec- 
tion with  the  variable  frequency  oscillator  at 
the  rear.  Panel  control  of  the  oscillator  is  ac- 
complished by  an  extension  shaft  and  a flexi- 
ble coupling.  At  the  left  of  the  chassis  is  the 
audio  system. 

Placement  of  the  main  tuning  dial  is  dic- 
tated by  the  height  of  the  three-gang  tuning 
capacitor  shaft  above  the  chassis  deck.  In  this 
case,  it  is  necessary  to  cut  a small  notch  in  the 
soft  aluminum  chassis  to  permit  the  dial  drive 
as.sembly  to  drop  low  enough  to  permit  the  dial 
shaft  to  align  with  the  capacitor  tuning  shaft. 

No  interstage  shields  are  required  below 
the  chassis.  Filter  choke  CHI  mounts  at  the 
rear  of  the  chassis,  directly  below  filament 
transformer  Ti.  The  100  ggfd.  filter  capacitor 
mounts  to  the  left  wall  of  the  chassis  below 
and  to  the  left  of  v-f-o  tuning  capacitor  C:. 
The  three  slug-tuned  coils  of  the  receiver  may 
be  seen  in  the  photograph.  Coil  Li  is  located 
behind  change-over  switch  Si.  Coil  L2  is  to- 
wards the  middle  of  the  chassis,  and  adjacent 
to  the  10  gfd.  audio  decoupling  capacitor.  Coil 
Ls  is  to  the  left  of  L2  and  between  the  r-f  tube 
and  mixer  tube  sockets. 


The  transmitter  oscillator  coil  is  placed  at 
the  left  rear  corner  of  the  chassis.  The  adjust- 
able slug  of  this  coil  is  visible  in  figure  39 
between  the  selenium  rectifiers  and  tuning  ca- 
pacitor Ci. 

To  facilitate  wiring,  four  phenolic  tie-point 
terminal  strips  (six  lug  type)  are  placed  be- 
neath the  chassis.  One  is  located  parallel  to 
the  edge,  just  to  the  left  of  the  6AQ5  buffer 
tube  socket.  A second  is  placed  parallel  to  the 
first,  to  the  right  of  the  same  socket.  The  latter 
strip  is  used  for  connections  to  the  r-f  circuitry 
of  the  receiver  section.  The  third  strip  is  placed 
parallel  to  the  rear  of  the  chassis  just  in  front 
of  the  i-f  strip  and  is  used  for  i-f  connections. 
The  last  strip  is  located  between  the  6AL5  tube 
socket  and  the  main  filter  capacitor.  The  noise 
limiter  and  a-v-c  components  are  mounted  to 
this  strip. 

To  secure  maximum  stability  it  is  necessary 
to  firmly  mount  the  oscillator  tuning  capacity 
to  the  chassis.  An  aluminum  block  measuring 
IH"  X 1"  was  therefore  cut  from  a section  of 
14 -inch  aluminum  stock  and  used  as  a support 
block  for  capacitor  C2.  The  capacitor  is  bolted 
to  the  chassis  by  long  6-32  machine  screws 
which  pass  through  the  block  and  are  firmly 
bolted  beneath  the  chassis. 


Figure  40 

UNDER-CHASSIS 
VIEW  OF 
TEN  METER 
TRANSCEIVER 

The  iransmHier  com> 
ponenis  are  along  the 
left  side  of  the  chassis. 
Receiver  components  are 
at  the  center,  with  r-f 
stage  nearest  the  panel. 
Filter  choke  is  mounted 
below  chassis  at  rear. 
Audio  stage  components 
are  grouped  in  upper 
right  corner  of  chassis. 
See  text  for  placement 
of  the  smaller  compon- 
ents* 


HANDBOOK 


28-Mc.  Transceiver  565 


3-WlRe  HOUSEHOLD 
WIRING  Cl  RCUIT 


"HOT"  LEAD  OF  TRANSCEIVER 
■ POWER  SUPPLY 


"GROUND"  LEAD  OF  TRANS- 
■“CEIVER  POWER  SUPPLY 


Figure  41 

''GROUNDED  NEUTRAL" 
WIRING  SYSTEM 

Most  dwellings  have  a three-wire  grounded 
neutral  system.  115  volts  a-c  can  be  ob- 
tained from  either  side  of  the  system  to 
ground.  Separate  ground  lead  on  trans- 
ceiver is  always  connected  before  line  plug 
is  energ'zed.  This  eliminates  usual  "ac-de'* 
shock  hazard. 


Transceiver  The  sequence  of  wiring  follows 
Wiring  much  the  same  pattern  as  that 

of  the  six  meter  transceiver 
described  in  section  27-6.  Small  components 
are  installed  between  tube  socket  pins,  or  be- 
tween socket  pins  and  adjacent  terminal  strip 
lugs.  Socket  ground  connections  ate  made  first 
and  then  the  filament  wiring  is  completed-  The 
power  supply  wiring  can  be  done  first. 
When  the  supply  is  completed  it  should  be 
tested.  A 2500  ohm,  25-watt  resistor  should 
be  attached  between  the  output  of  the  supply 
and  ground.  A full  250  volts  should  be  devel- 
oped across  this  resistor  with  115  volts  a-c 
applied  to  the  power  supply. 

Next,  the  i-f  and  second  detector  section  of 
the  receiver  should  be  wired.  Short,  direct 
leads  and  the  use  of  the  adjacent  terminal 
strips  will  reduce  crowding  in  this  area.  The 
audio  stages  should  now  be  wired,  along  with 
switch  segments  SIB,  SID,  and  SIE.  When 
this  is  completed,  the  i-f  section  and  audio  of 
the  transceiver  may  be  tested.  A test  speaker 
should  be  connected  to  jack  Js,  and  a 1.5  Me. 
tone  modulated  signal  from  a test  oscillator  is 
loosely  coupled  to  the  plate  pin  (#6)  of  the 
6U8  socket.  The  top  and  bottom  slugs  of  the 
i-f  transformers  are  now  adjusted  for  maxi- 
mum output  signal.  As  the  stages  approach 
alignment,  be  sure  to  decouple  the  signal  gen- 
erator to  prevent  overloading. 

After  this  portion  of  the  transceiver  has 
been  tested,  the  r-f  circuits  of  the  receiver 
should  be  wired.  Coils  Lj,  La,  and  La  are  wound 
and  the  coil  forms  clipped  in  place.  The  front 
section  of  tuning  capacitor  CIA  is  employed 
for  the  grid  circuit  of  the  r-f  stage,  the  middle 
section  is  for  the  interstage  r-f  circuit,  and  the 
tear  section  is  for  the  oscillator  section  of  the 
6U8  tube.  Connections  between  the  coils,  the 


tuning  capacitor,  and  the  tube  sockets  are  made 
with  # 16  tinned  solid  wire.  In  particular,  care 
should  be  taken  to  prevent  movement  of  the 
leads  and  components  in  the  oscillator  section 
as  any  minute  vibration  of  this  part  of  the 
circuit  would  lead  to  receiver  instability. 

Sufficient  coupling  exists  between  the  wir- 
ing of  the  oscillator  and  mixer  circuits  to  pro- 
vide the  proper  level  of  injection,  and  no 
coupling  capacitor  is  required.  The  final  wir- 
ing step  is  to  wind  the  antenna  coil  on  form 
Li.  One  end  is  attached  to  a nearby  socket 
ground  lug  and  the  other  end  goes  to  the 
"receive”  contact  of  transfer  switch  SIA. 

When  the  receiver  is  completed,  a 28.5  Me. 
may  be  injected  into  the  antenna  circuit  and 
the  main  tuning  dial  set  near  maximum  ca- 
pacity. The  slug  of  receiver  oscillator  coil  Ls 
is  now  adjusted  until  the  signal  is  received  near 
a dial  setting  of  thirty  degrees.  After  this, 
coil  L)  and  coil  Ls  are  peaked  for  maximum 
signal  strength.  Dial  calibration  should  wait 
until  the  transmitter  section  is  completed. 

Completion  of  the  The  oscillator  coil  should 
Tronsmitter  Section  be  wound  and  mounted 
in  position.  The  lead 
from  coil  Lj  to  tuning  capacitor  G is  made  of 
# 14  solid  copper  and  passes  through  a M-inch 
hole  in  the  chassis.  All  oscillator  components 
are  firmly  mounted  to  reduce  vibration  to  a 
minimum.  Coils  Ls  and  D are  wound  and 
snapped  into  position  as  the  buffer  stage  is 
being  wired.  The  p-a  plate  coil  (Li)  is  mount- 
ed above  chassis  between  the  6AQ5  amplifier 
tube  and  the  front  panel.  The  plate  (or  "hot”) 
end  of  the  coil  is  supported  on  a Vi -inch  cer- 
amic insulator  and  the  coil  lead  passes  through 
a V4-iuch  hole  in  the  chassis  to  the  stator  of 
tuning  capacitor  Cs,  mounted  directly  below 
the  coil.  The  opposite  end  of  coil  L?  is  attached 
to  a polystyrene  feed-through  insulator.  Neu- 
tralizing capacitor  Cs  is  affixed  at  one  end  to 
a terminal  of  coil  form  Ls.  The  other  end  at- 
taches to  the  plate  t-f  choke,  as  shown  in  the 
under-chassis  photograph.  The  output  lead 
from  transfer  switch  SIA  to  the  antenna  re- 
ceptacle on  the  teat  of  the  chassis  is  made  from 
a short  length  of  coaxial  cable.  When  the 
wiring  is  completed  all  connections  should 
be  checked  and  the  chassis  thoroughly  cleaned 
of  solder  bits,  pieces  of  wire,  flux,  etc. 

Transmitter  The  6BJ6  oscillator  tube  should 
Adjustment  be  placed  in  the  socket,  along 
with  the  OA2  voltage  regulator. 
Transmitter  frequency  control  capacitor  C2  is 
iet  near  full  setting  and  the  slug  of  oscillator 


566  Receivers  and  Transceivers 


coil  L,  is  adjusted  so  as  to  place  the  oscillator 
frequency  on  7.1  Me.  The  fourth  harmonic  of 
the  oscillator  will  then  be  on  28.4  Me.  At 
minimum  tuning  capacitor  setting  the  upper 
frequency  limit  of  the  transmitter  will  be 
slightly  above  29.7  Me.  The  6AQ5  tubes  are 
now  placed  in  their  sockets,  and  a 0.5  d-c  mil- 
liammeter  is  temporarily  connected  across 
meter  shunt  Rj  in  the  grid  circuit  of  the  6AQ5 
final  amplifier  stage.  The  screen  lead  (pin  6) 
of  the  6AQ5  amplifier  tube  is  temporarily 
opened.  Transfer  switch  Si  is  placed  in  the 
"transmit”  position  and  the  v-f-o  tuning  ca- 
pacitor C2  is  set  to  29  Me.  The  transmitter  is 
energized  and  the  slugs  of  oscillator  coil  Lb 
and  buffer  coil  Ls  are  adjusted  for  maximum 
grid  drive  to  the  amplifier  (about  two  milli- 
amperes).  Amplifier  loading  capacitor  C*  is 
set  at  full  capacity,  and  tuning  capacitor  G is 
tuned  through  its  range  while  carefully  noting 
the  grid  current  reading.  The  reading  will  show 
an  abrupt  kick  as  G is  tuned  through  reson- 
ance. Next,  the  setting  of  neutralizing  capaci- 
tor G should  be  slowly  varied  by  means  of  a 
fibre-blade  screwdriver  so  as  to  minimize  the 
kick  of  grid  current.  After  each  movement  of 
G,  the  slug  of  coil  Le  should  be  reset  for  max- 
imum grid  current.  A setting  of  G can  readily 
be  found  that  will  provide  a minimum  value 
of  grid  current  change  as  capacitor  G is  tuned 
through  resonance.  The  last  step  is  to  resolder 
the  screen  lead  to  pin  6 of  the  6AQ5  amplifier 
tube  socket. 


Transceiver  The  transmitter  may  now  be 
Operation  attached  to  a suitable  antenna 
or  dummy  load  (see  section 
27-6)  for  an  operative  check-out.  Receiver 
calibration  should  be  rechecked,  and  the  dial 
may  be  calibrated.  The  transfer  switch  should 
be  set  to  the  transmitting  position  and  the 
transceiver  loaded  into  the  antenna  system.  The 
meter  now  indicates  the  r-f  voltage  rectified  by 
the  9006  diode  attached  to  the  output  of  the 
pi-network  system.  Tuning  (G)  and  loading 
(Cl)  should  be  adjusted  to  provide  a maxi- 
mum meter  reading,  with  Cs  always  being  set 
last  for  the  final  "touch  up”  adjustment.  Maxi- 
mum capacity  setting  of  Ci  provides  ’minimum 
loading  and  vice-versa. 

It  may  be  necessary  when  working  into  low 
impedance  loads,  or  transmission  lines  having 
a high  value  of  SWR  to  parallel  loading  ca- 
pacitor Cl  with  an  auxiliary  100  ggfd.  mica 
capacitor  to  obtain  optimum  loading.  This  will 
have  to  be  determined  by  experiment. 

A mobile  carbon  microphone  may  now  be 
plugged  into  jack  J2.  An  indication  of  the  mod- 
ulation level  can  be  obtained  by  noting  the 
increase  in  brilliance  of  the  lamps  of  the  dum- 
my antenna.  Overloading  will  tend  to  cause 
downward  modulation.  The  degree  of  modula- 
tion may  be  varied  by  moving  the  microphone 
away  from  the  lips.  Using  a telephone-type 
V-1  unit,  optimum  modulation  occurs  when 
speaking  in  a normal  tone  about  three  inches 
from  the  microphone.  The  transceiver  output 
is  about  six  watts  fully  modulated  with  a power 
amplifier  input  of  ten  watts. 


CHAPTER  TWENTY-EIGHT 


The  exciter  is  the  "heart”  of  the  amateur 
transmitter.  Various  forms  of  amplifiers  and 
power  supplies  may  be  used  in  conjunction 
with  basic  exciters  to  form  transmitters  which 
will  suit  almost  any  need.  Of  great  interest 
today  ate  simple  SSB  exciters  for  fixed  and 
mobile  operation.  These  may  be  used  "as-is” 
or  with  a small  linear  amplifier  for  mobile 
work,  or  may  be  combined  with  a high  power 
linear  amplifier  for  fixed  station  operation. 
Also  occupying  a position  of  importance  is 
the  "package”  VHP  station  capable  of  opera- 
tion on  one  or  more  of  the  VHP  amateur 
bands. 

Several  different  types  of  equipment  de- 
signed to  meet  a wide  range  of  needs  are  des- 
cribed in  this  chapter.  There  are  two  different 
sidband  exciters  for  fixed/mobile  service  and 
a de-luxe  VHP  station  capable  of  outstanding 
performance  on  two  VHP  amateur  bands.  Also 
shown  is  a high  stability  variable  frequency 
oscillator  unit  designed  for  operation  in  the 
high  frequency  DX  bands.  To  the  amateur 
who  is  interested  in  the  construction  phase  of 


his  hobby,  these  and  other  units  shown  in  later 
chapters  should  offer  interesting  ideas  which 
might  well  fit  in  with  the  design  of  his  basic 
transmitting  equipment. 

As  in  the  previous  chapter,  the  component 
nomenclamre  and  color  codes  outlined  in  fig- 
ures 1,  2,  and  3 of  chapter  27  are  used  in  this 
chapter  unless  otherwise  noted. 

28-1  SSB  Exciter  for 
Fixed  or  Mobile  Use 

The  simplest  and  most  economical  method 
of  generating  a single  sideband  signal  is  to 
employ  a phasing-type  transmitter  of  the  form 
outlined  in  chapter  17  of  this  Handbook.  If 
the  r-f  phasing  system  operates  on  the  carrier 
frequency  of  the  transmitter  the  complex  fre- 
quency conversion  circuits  may  be  omitted  and 
the  complete  exciter  becomes  inexpensive  to 
build  and  simple  to  place  in  operation. 

A SSB  exciter  suitable  for  fixed  or  mobile 
operation  is  shown  in  figures  1 and  4.  The 
exciter  delivers  a 3 watt  peak  power  signal, 


567 


568  Low  Power  Transmitters 


THE  RADIO 


sufficient  to  drive  a high  power  tetrode  linear 
amplifier  to  a kilowatt  level.  Operation  is  con- 
fined to  the  80  meter  band,  although  operation 
on  the  higher  frequency  bands  is  possible  with 
a change  of  coils,  phasing  network,  and  crystal. 
The  r-f  phasing  circuits  are  balanced  at  some 
frequency  within  the  amateur  phone  band  and 
will . retain  a good  degree  of  balance  over  a 
frequency  range  of  plus  or  minus  fifty  kilo- 
cycles of  the  adjustment  frequency. 

Circuit  The  circuit  of  the  single  side- 

Description  band  exciter  is  shown  in  figures 

2 and  3.  A 12AU7  is  employed 
as  a crystal  oscillator  stage  and  buffer-amplifier. 
The  first  section  of  the  double  triode  is  used 
in  a Pierce  oscillator  circuit  with  the  crystal 
connected  between  the  grid  and  plate  of  the 
tube.  The  oscillator  operates  directly  on  the 
chosen  SSB  frequency  in  the  80  meter  band. 
The  frequency  of  oscillation  may  be  varied  over 
a range  of  two  hundred  cycles  or  so  by  the  50 
ggfd  trimming  capacitor  permitting  the  trans- 
mitter to  be  "zeroed  in"  on  a particular  SSB 
channel.  The  second  section  of  the  12AU7 
serves  as  an  isolation  amplifier  with  the  plate 
circuit  tuned  to  the  operating  frequency. 
Changes  in  the  input  impedance  of  the  diode 
modulator  stage  under  operating  conditions 
would  cause  frequency  shift  of  the  oscillator 
stage  if  direct  coupling  between  these  circuits 
was  used.  The  isolation  afforded  by  the  buffer 
stage  effectively  prevents  frequency  pulling  of 
the  oscillator  stage  during  modulation. 

The  output  of  the  buffer  stage  is  link 
coupled  to  a simple  90°  r-f  phase  shift  net- 
work wherein  the  audio  signal  from  the  audio 
phasing  amplifier  is  combined  with  the  r-f 
signals.  The  network  is  of  the  R-C  type  made  up 
of  50  ohm  non-inductive  resistors  and  800 
ggfd.  capacitors  in  a bridge  configuration.  The 
reactance  of  the  capacitors  is  very  close  to  50 


IEAU7  4-1N81  6CL6 


CONTROL 

Figure  2 

BLOCK  DIAGRAM  OF  PHASING-TYPE 
SIDEBAND  EXCITER 
Only  five  tubes  are  required  in  this  simple 
phasing-type  sideband  exciter.  Audio  system 
has  sufficient  gain  to  operate  from  crystal 
microphone. 


ohms  in  the  center  of  the  80  meter  phone 
band.  Bridge  balance  and  carrier  elimination 
is  achieved  by  adjustment  of  the  variable  po- 
tentiometers (Rs,  R9,  Rio)  in  the  bridge  cir- 
cuit. Four  selenium  diodes  are  used  in  the 
modulator  circuit.  A 1N82  multiple  diode  as- 
sembly may  be  employed,  or  four  1N81  diodes 
whose  front/back  ratios  are  equal  may  be  used. 

The  output  of  the  balanced  modulator  net- 
work is  coupled  by  a low  impedance  link  to 
the  grid  circuit  of  a neutralized  6CL6  linear 
amplifier.  This  stage  operates  class  ABi  with 
cathode  bias,  delivering  a 3 watt  peak  SSB 
signal  to  a low  impedance  load  circuit.  All 
circuits  are  tuned  by  adjustable  slug-tuned 
coils  ( Li,  L:,  and  Ls ) . 

A cascade  12AU7  and  a 6C4  comprise  the 
speech  amplifier  which  may  be  driven  from  a 
crystal  microphone.  The  output  of  the  6C4 
is  transformer  coupled  into  the  audio  phase 
shift  network  PS-1.  An  interstage  transformer 
is  connected  backwards  for  Ti,  providing  a 
relatively  low  impedance  secondary  winding 
delivering  two  equal  audio  voltages  that  are 
180°  out  of  phase.  These  voltages  are  ap- 
plied to  the  special  audio  network  whose  out- 
put circuits  drive  separate  triode  sections  of 
the  12AU7  audio  phasing  amplifier.  The 
12AU7  tube  functions  as  a dual  cathode  fol- 


Figure  1 

THE  GLOVE  COMPARTMENT 
SIDEBAND  EXCITER 

This  miniature  phasing-type  SSB  exciter  may 
be  mounted  in  the  glove  compartment  of  a 
cor  with  room  to  spare!  Using  only  four  tubes 
the  exciter  delivers  a 3 watt  peak  sideband 
signal  capable  of  driving  a kilowatt  tetrode 
amplifier  to  full  output.  The  I2AU7  and 
6CL6  r-f  tubes  are  at  the  right  end  of  the 
chassis,  along  with  the  crystal.  Major  tuning 
controls  are  on  the  right  end  of  the  chassis. 
Satire  unit  is  bolted  to  the  roof  of  the  glove 
compartment. 


HANDBOOK 


S.S.B.  Exciter  569 


lower  which  provides  the  necessary  low  im- 
pedance required  to  match  the  r-f  phasing  net- 
work. 

Sideband  switching  is  accomplished  by  re- 
versal of  polarity  of  the  audio  channels  by 
switch  Si.  The  diode  modulator  may  be  un- 
balanced to  pass  a carrier  for  tune-up  purposes 


by  throwing  carrier  insertion  switch  S=  to  the 
right  and  biasing  the  diode  modulators  by 
means  of  carrier  insertion  control  Ri. 

Power  requirements  for  the  transmitter  are 
300  volts  at  a current  drain  of  100  milliam- 
peres,  and  6.3  volts  d-c  at  1.85  amperes.  When 
the  filaments  are  connected  for  12.6  volt  op- 


6 VOLT 

12AU7  4.5  A 9 


12AU7  k.5A9  1 
AUDIO  “ ^ 


PS-1 

(S££  F/GU^f£  7 ) 


NOTE 

■K-  = MATCHED  COMPONENTS  {l  % OR 
LESS).  EXACT  VALUE  CRITICAL 
ONLY  IN  THAT  IT  SHOULD  MATCH 
THE  MATING  UNIT  CLOSELY. 


TO  Pi 


Figure  3 


SCHEMATIC  DIAGRAM  OF  FIXED/MOBILE  SSB  EXCITER 


Cl,  Ct,  Cj,  Ci — See  figure  7 

Cs — 70  /xyufd  variable  ceramic,  Centrafab  827B. 

R],  Ri,  Ra,  Ri — See  figure  7 
Ro — SOOK  linear  taper  potentiometer 
Rs,  Rio— 100  ohm  potentiometer 
Rt — 25K  linear  taper  potentiometer 
Rs,  Rs — IK  potentiometer 

PS-1 — Phase  shift  network  package.  See  figure  7 
Li,  Li,  Ls — See  figure  8 

Ti — Interstage  audio  transformer  (1:3).  Staneor  A-S3C  used  backwards. 

RFC — Miniature  2.5  mh.  choke.  Millen  J300-2500 
X — 80  meter  crystal  (3800  - 4000  kc.) 

PC — 3 turns  # 78  e.  wire  around  50  ohm,  0.5  wott  composition  resistor. 
Di-Di — 1N81  diodes;  see  text. 


570  Low  Power  Transmitters 


THE  RADIO 


TOP  VIEW  OF  SSB  EXCITER 

The  top  plate  is  removable  for  easy  access  to 
internal  wiring  and  adjustments.  Microphone 
jack  and  coaxial  antenna  receptacle  are  visi- 
ble at  left  of  chassis.  Four  control  potentio- 
meters are  along  bottom  of  box  (see  text). 


eration  current  the  current  drain  is  1.25  am- 
peres. 

Transmitter  The  transmitter  is  constructed 

Construction  upon  an  aluminum  box-chas- 

ond  Wiring  sis  measuring  10"  x IVl"  x 
2V2"  in  size.  Placement  of 
the  major  components  may  be  seen  in  figure  5. 
The  80  meter  crystal  and  12AU7  oscilla- 
tor/buffer tube  are  positioned  at  the  left  end 
of  the  chassis,  and  the  6CL6  linear  amplifier 
is  centered  on  the  chassis.  The  balanced  modu- 
lator is  mounted  on  a phenolic  terminal  board 
on  the  opposite  side  of  the  box.  Oscillator  tun- 
ing controls  are  placed  on  the  left  side  of  the 
box-chassis,  along  with  the  sideband  selector 
switch  Si,  the  carrier  insertion  switch  S2,  and 
the  audio  gain  control  Rs.  Along  the  "bottom” 
of  the  chassis  are  placed  (left  to  tight)  the 
modulator  balance  potentiometers  R»,  R»,  and 
Rio  and  the  audio  balance  control  R«.  The  car- 
rier insertion  potentiometer  R?  is  mounted  in- 
side the  chassis.  The  three  audio  stages  are 
mounted  on  the  right-hand  section  of  the 
chassis. 

The  components  of  the  audio  phasing  net- 
work PS-1  are  mounted  on  a small  phenolic 
terminal  board  that  may  be  observed  in  the 
lower  right  corner  of  the  chassis.  This  network 
should  be  wired  and  tested  before  it  is  placed 
in  the  transmitter. 

The  Audio  The  complete  network 

Phasing  Network  may  be  purchased  as  a 
finished  item  (Millen 
75012)  or  it  may  easily  be  home  built  if 
an  audio  oscillator  and  oscilloscope  are  avail- 
able for  testing  purposes. 

The  circuit  of  the  audio  network  is  shown 


Figure  5 

UNDER-CHASSIS  VIEW  OF  SSB 
EXCITER 

Placement  of  the  main  components  may  be 
seen  in  this  inside  view  of  the  chassis.  R~f 
circuitry  is  at  the  left  with  home-made  audio 
tdtasing  network  at  center,  right.  R-f  modu- 
lator potentiometers  and  germanium  diodes 
in  lower  left  area  of  the  chassis  with  audio 
section  at  right.  The  6CL6  linear  amplifier 
plate  coil  is  in  the  foreground  of  the  center 
of  the  chassis. 


in  figure  7 A.  The  mounting  base  may  be  thin 
phenolic  or  any  insulating  material.  The  base 
holds  four  precision  resistors  and  four  fixed 
mica  capacitors  padded  with  four  adjustable 
mica  trimmer  capacitors.  If  desired,  a series 
of  fixed  capacitors  may  be  measured  on  a 
capacity  bridge  and  hand  picked  units  chosen 
to  replace  the  variable  capacitors  after  final 
adjustments  have  been  completed.  This  was 
done  with  the  network  shown  in  the  photo- 
graphs- 

The  two  100  K series  resistors  used  in  the 
network  are  Continental  Noheloy  1%  tolerance 
precision  resistors.  The  133.3  K resistors,  how- 
ever, were  made  by  taking  two  150  K pre- 
cision Continental  Nobeloy  resistors  and  paral- 
leling each  of  them  with  a one-half  watt  1.2 
megohm  (plus  or  minus  10%  tolerance)  resis- 
tor. Careful  selection  of  the  1.2  M units  per- 
mits close  adjustment  to  the  desired  target 
value  of  133.3  K.  A convenient  way  to  mount 
the  1.2  M resistors  is  to  slip  them  inside  the 
hollow  body  of  the  precision  resistors.  The 
dashed  connections  should  be  omitted  initially, 
since  the  alignment  procedure  described  below 
presumes  that  these  connections  will  be  made 
at  the  proper  time  only. 

Adjustment  of  the  If  a manufactured  net- 
Audio  Network  work  is  employed  no  ad- 

justments are  required. 
However,  the  home-made  network  must  be 
aligned  before  it  is  placed  in  the  transmitter 


HANDBOOK 


S.S.B.  Exciter  571 


chassis.  The  network  resistors  should  bear  the 
ratio  of  133-3  to  100.0,  that  is,  4 to  3 as  close- 
ly as  can  be  determined.  If  in  doubt  as  to  the 
ratio  of  the  resistors  you  use,  double-check 
their  values  on  an  accurate  bridge.  The  ad- 
justment of  the  phase  shift  network  now  con- 
sists only  of  setting  the  four  capacitors  to  their 
proper  values. 

An  audio  oscillator  capable  of  operation 
from  225  to  2750  cycles  per  second  (with 
good  waveform)  is  required,  plus  an  oscillo- 
scope. The  oscillator  should  be  carefully  cali- 
brated by  the  method  described  later.  Connect 
the  output  of  the  audio  oscillator  through  a 
stepdown  transformer  (Ti  will  serve  nicely) 
to  a 1 K potentiometer  (use  Ri)  with  the  arm 
grounded. 

Adjust  the  arm  position  so  that  equal  and 
opposite  voltages  appear  on  each  half  of  the 
potentiometer.  A steady  audio  frequency  signal 
of  any  convenient  frequency  may  be  used  with 
the  oscilloscope  acting  as  the  voltmeter  for 
this  job.  Swing  the  vertical  deflection  lead  of 
the  ’scope  from  one  end  of  the  potentiometer 
to  the  other  and  adjust  the  arm  to  obtain  equal 
voltages.  Hook  up  a temporary  double  cathode 
follower  using  a 12AT7  with  500  ohms  from 
each  cathode  to  ground  and  connect  as  shown 
in  figure  7B.  (It  will  be  convenient  to  pro- 
vide leads  M,  N,  and  1 and  2 with  clips  at 
the  ends  to  facilitate  checking).  Cathode  pins 
8 and  3 of  the  12AT7  should  connect  to  the 
H and  V deflection  amplifiers  in  the  oscillo- 
scope, and  the  oscilloscope  “ground"  connec- 
tion should  be  made  to  the  common  return  of 
the  cathode  follower  amplifier. 

First  connect  lead  M to  terminal  A on  PS-1 
and  lead  N to  terminal  A'.  Connect  leads  1 
and  2 to  terminal  M.  (Note  that  the  dashed 
connections  are  missing  at  this  stage  of  ad- 


Figure  6 

SIMPLE  "VOX"  CIRCUIT  FOR  ANY 
SSB  EXCITER 

Capacitor  C and  resistor  R determine  the  "de- 
lay time"  of  the  voice  operated  relay  circuit. 
Operation  point  is  set  by  potentiometer  Ri. 
Relay  RYi  has  a high  resistance  d.c.  coil 
(Automatic  Electric  Co.  type  R-45L  with 
10,000  ohm  coil  may  be  used). 


justment).  Adjust  the  horizontal  and  vertical 
gain  controls  of  the  oscilloscope  to  produce  a 
line  about  1 1/2  inches  long  slanted  at  45  de- 
grees when  the  oscillator  is  set  to  a frequency 
of  490  c.p.s.  (an  exact  method  of  setting  fre- 
quency will  be  described  later).  If  the  oscillo- 
scope has  negligible  internal  phase  shift  the 
display  will  be  a straight  line  instead  of  a nar- 
row slanting  ellipse.  If  the  latter  display  ap- 
pears, it  is  necessary  to  correct  the  oscilloscope 
phase  shift  externally  by  using  an  adjustable 
series  resistance  (a  100  K potentiometer) 
mounted  at  either  the  vertical  or  horizontal 
input  terminal,  depending  on  what  correction 
is  necessary. 

The  objective  here  is  to  obtain  a single 
straight  line  pattern  on  the  ’scope  at  an  audio 
frequency  of  490  c.p.s.  In  some  cases  a series 
capacitor  may  be  needed  in  conjunction  with 
the  potentiometer  to  provide  the  necessary  cor- 
rection. Try  values  from  .05  gfd  to  .0005  /xfd. 
'When  the  required  slanted  line  is  obtained. 


PHASE  SHIFT  NETWORK  PS-1 


Tl.itS 

STANCOR  AS3-C 


AUDIO  OSCILLATOR 


PHASE 

SHIFT 

NETWORK 

PS-1 


m 


* TO  VERTICAL  DEFLECTION  PLATE 


$500 


► TO  HORIZONTAL  DEFLECTION  PLATE 


Figure  7 

AUDIO  PHASE-SHIFT  NETWORK  AND  TEST  LAYOUT 


572  Low  Power  Transmitters 


THE  RADIO 


NOTE  : rivo  parallel  pieces  of  hookup  wire  are  wound 

SIMULTANEOUSLY  OVER  PRIMARY  WIN0IN6  TO  MAKE 
BIFILAR  SECONDARY.  BOTTOM  END  OF  ONE  WINDINS 
AND  TOP  END  OF  OPPOSITE  WINDING  ARE  GROUNDED 
TO  MOUNTING  LUG  ON  COIL  FORM. 

Figure  8 

BIFILAR  COIL  FOR  SSB 
TRANSMITTER  (Li) 

Coils  Li  and  L3  have  4-iurn  link  coils  over 
bottom  end  instead  of  the  bifilar  winding. 


shift  lead  1 from  terminal  A to  terminal  B 
on  the  phase  network  PS-1.  Adjust  the  trimmer 
capacitor  Ci  (figure  3)  to  obtain  a circle  on 
the  oscilloscope.  It  will  be  noted  as  this  ad- 
justment is  made  the  display  will  shift  from 
an  ellipse  "leaning"  to  one  side  through  a 
circle  or  ellipse  (with  axes  parallel  to  the  de- 
flection axes)  to  an  ellipse  which  "leans”  the 
other  way.  If  desired,  the  appropriate  gain 
control  on  the  oscilloscope  may  be  varied  so 
that  a circle  instead  of  an  "erect”  ellipse  is 
obtained  at  the  point  of  correct  adjustment. 
After  checking  the  gain  control  on  the  ’scope, 
recheck  and  correct  (if  necessary)  the  phase 
shift  in  the  oscilloscope  by  moving  lead  1 
back  to  terminal  A,  and  then  repeat  the  setting 
of  capacitor  Ci  with  lead  1 back  on  terminal  B. 

In  general,  always  make  certain  that  the  os- 
cilloscope is  used  in  a phase-corrected  manner. 
As  a double  check  (if  the  deflection  plates  in 
the  oscilloscope  are  skewed,  for  instance)  con- 
nect lead  2 to  terminal  A'.  If  the  circle  changes 
to  a slanting  ellipse,  readjust  Ci  to  produce  an 
ellipse  "half-way”  between  the  ellipse  (ob- 
tained by  switching  lead  2)  and  a circle. 
Changing  lead  2 from  A'  to  A and  back 
again  should  give  equal  and  opposite  skew 
to  the  display  when  capacitor  Ci  is  set  correctly. 
Failure  to  get  symmetrical  ellipses  (egg-shaped, 
or  other  display)  is  due  to  distortion  in  the 
’scope,  the  oscillator,  the  transformer,  or  the 
cathode  follower.  Conduct  the  test  at  the  low- 
est signal  level  possible  to  avoid  distortion. 

Next,  connect  leads  M and  N to  terminals  E 
and  E',  respectively.  Connect  leads  1 and  2 to 


E,  set  the  oscillator  frequency  to  I960  c.p.s., 
correct  oscilloscope  phase  shift  as  before,  and 
move  lead  1 to  terminal  G.  Adjust  capacitor 
Ci  in  PS-1  for  a circle  as  was  done  for  Ci, 
using  the  precautions  outlined  for  that  case. 

Now  connect  lead  M to  terminal  D,  and 
lead  N to  terminal  F.  Connect  leads  1 and  2 
to  terminal  D,  set  the  oscillator  frequency  at 
1307  c.p.s.,  correct  oscilloscope  phase  shift 
as  before,  and  move  lead  1 to  the  junction  of 
Ra  and  Ca.  Adjust  capacitor  Ca  for  a circle  on 
the  oscilloscope,  as  before. 

Repeat  the  above  procedure  for  the  remain- 
ing R-C  pair,  R,  and  Ca.  Use  terminals  D and 
C at  this  time  and  set  the  oscillator  to  326.7 
c.p.s.  This  completes  the  tests  except  for  a final 
check  to  all  adjustments.  Connect  A to  A',  E 
to  E , B to  C,  F to  G,  and  A to  E.  The  net- 
work PS-1  is  now  completely  wired. 

Checking  rte  If  the  oscilloscope  did  not 
Audio  Network  require  changes  in  external 
compensation  over  the  four 
frequencies  used,  an  over-all  frequency  check 
may  now  be  easily  made  on  the  phase  shift  net- 
work. To  do  this,  connect  lead  1 to  point 
B-C,  lead  2 to  point  F-G,  lead  M to  point  A-A'- 
E-E',  and  lead  N to  point  D.  Now  shift  the 
arm  of  the  potentiometer  towards  M until  a 
circle  appears  on  the  oscilloscope  screen  at  a 
frequency  of  250  c.p.s.  Then,  as  the  oscillator 
frequency  is  varied  from  250  c.p.s.  to  2500 
c.p.s.,  this  circle  will  wobble  a little  from  one 
side  to  the  other,  passing  through  a perfect 
circular  display  at  440,  1225,  and  2500  c.p.s. 

The  audio  band  over  which  the  wobble  in- 
dicates a plus  or  minus  1.3  degree  deviation 
from  true  90  degrees  is  225  to  2750  c.p.s.,  or 
12  to  1 in  range.  This  means  that  when  other 
circuits  are  properly  adjusted,  a sideband  sup- 
pression ratio  of  39  decibels  is  possible  at  the 
worst  points  within  this  range.  The  average 
suppression  ratio  will  be  about  45  decibels. 
Proper  phase-shift  network  operation  is  neces- 
sary to  obtain  this  class  of  performance,  so 
the  adjustment  procedures  have  been  outlined 
in  detail  as  an  aid  toward  this  goal.  The  phase 
shift  network  should  never  require  readjust- 
ment, so  that  when  you  are  satisfied  with  the 
adjustment  you  may  seal  the  trimmers  with 
cement. 

Audio  Oscillafor  It  will  be  noted  that  the 
Calibration  frequency  ratios  are  such 

that  the  12th  harmonic 
of  326.7  c.p.s.,  the  8th  harmonic  of  490 
c.p.s.  and  the  3rd  harmonic  of  1306.7  c.p.s. 
are  all  the  same  as  the  2nd  harmonic  of 


HANDBOOK 


S.S.B.  Exciter  573 


I960  c.p.s..,  namely  3920  c.p.s.  Thus,  if  a 
stable  source  of  3920  c.p.s.  frequency  (such 
as  a thoroughly  warmed  audio  oscillator) 
is  used  as  a reference,  the  frequency  of  the 
test  oscillator  can  be  set  very  closely  to  one- 
half,  one-third,  etc.  of  the  reference  frequency 
if  both  oscillators  feed  an  oscilloscope  and  the 
resulting  Lissajous  figures  observed,  as  dis- 
cussed in  chapter  9 of  this  Handbook. 

Use  of  a calibrating  frequency  in  this  man- 
ner assures  that  the  frequency  ratios  used  are 
correct,  even  though  the  exact  frequencies  used 
are  unknown.  The  frequency  ratios  (just  as 
the  resistance  ratio  previously  mentioned)  are 
far  more  important  that  the  actual  values  of 
frequency  (or  resistance)  used. 

Transmitter  The  r-f  and  audio  sections  of 
Wiring  the  unit  should  be  wired  before 

the  phase  network  is  placed 
within  the  chassis.  Socket  ground  connections 
and  filament  wiring  are  done  first.  The  oscilla- 
tor section  should  be  wired  and  tested  first. 
The  bifilat  secondary  winding  of  coil  Li  may 
be  made  of  two  lengths  of  insulated  hookup 
wire  wound  over  the  coil  as  shown  in  figure  8. 

The  germanium  diodes  in  the  modulator 
section  deserve  special  care  in  handling.  Do  not 
bend  the  leads  close  to  the  diode  unit  itself. 
The  diodes  ate  mounted  by  means  of  their  leads 
between  the  terminals  of  potentiometers  Rs  and 
Rs  and  adjacent  tie-point  terminals.  Protect 
the  diodes  from  excessive  heat  while  soldering 
by  holding  the  lead  with  pliers  between  the 
body  of  the  diode  and  the  point  where  the  sol- 
dering takes  place.  Further,  use  only  as  much 
heat  as  is  necessary  to  make  a good  joint. 

The  50  ohm  composition  resistors  used  in 
the  modulator  bridge  circuit  should  be  a 
matched  pair,  measured  on  an  ohmmeter  or 
Wheatstone  bridge.  In  the  same  fashion  the 
bridge  capacitors  should  be  matched  as  closely 
as  possible  to  the  target  value  of  800  /x/xfd. 

Both  coils  Li  and  Ls  of  the  linear  amplifier 
are  mounted  beneath  the  chassis  on  either  side 
of  the  6CL6  tube  socket.  To  reduce  coupling 
between  the  coils,  a shield  is  placed  over  coil 
Li.  The  shield  may  be  made  from  part  of  an 
old  i-f  transformer  can  measuring  approxi- 
mately 1"  X 1”  X 2"  in  size. 

When  the  transmitter  wiring  is  completed, 
the  audio  phase  network  PS-1  may  be  placed 
in  the  transmitter  chassis  and  wired  in  place. 
Tronsmitter  After  the  wiring  has  been  com- 
Adjustment  pleted  and  checked  the  trans- 
mitter is  ready  for  adjustment. 
Insert  the  12AU7  oscillator/buffer  tube  in  the 
socket,  and  plug  in  a crystal  in  the  range  of 


3.8  - 4.0  Me.  Apply  plate  power  and  adjust 
the  slug  of  coil  Li  for  maximum  signal 
strength  of  the  oscillator  in  a nearby  receiver. 
The  oscillator  should  start  promptly  each  time 
plate  voltage  is  applied  to  the  unit.  Frequency 
of  oscillation  may  be  varied  over  a small  range 
by  changing  the  setting  of  the  crystal  tuning 
capacitor. 

The  next  step  is  to  neutralize  the  6CL6 
linear  amplifier.  It  is  necessary  to  bypass  the 
diode  modulator  to  do  this.  Temporarily  dis- 
connect one  lead  of  the  bifilar  link  winding 
on  Li,  and  also  disconnect  the  lead  between  po- 
tentiometer Rio  and  amplifier  grid  coil  Li.  Now, 
connect  the  free  winding  of  the  bifilar  link  of 
Li  to  the  ungrounded  end  of  the  link  of  Li. 
This  will  apply  the  oscillator  signal  directly 
to  the  linear  amplifier.  Temporarily  load  the 
6CL6  stage  with  one  or  two  flashlight  bulbs 
(6.3  volt,  150  ma.  — brown  bead)  connected 
in  parallel  at  the  output  jack  Ji.  The  slugs  of 
coils  L;  and  Ls  should  be  adjusted  for  maxi- 
mum amplifier  output.  The  12AU7  oscillator 
tube  should  now  be  removed  from  the  socket, 
and  unless  the  setting  of  neutralizing  capaci- 
tor Cs  happens  by  chance  to  be  correct,  the 
flash  lamp  load  will  continue  to  give  some  in- 
dication of  output,  showing  that  the  6CL6 
amplifier  stage  is  oscillating.  Neutralizing  ca- 
pacitor Cs  should  be  adjusted  with  the  aid  of 
a fibre  screwdriver  until  the  bulb  goes  out. 
Readjust  Li  and  Ls  for  maximum  bulb  indica- 
tion as  the  setting  of  capacitor  Ci  is  varied. 
A setting  of  Cs  will  be  found  at  which  oscilla- 
tion will  not  occur,  regardless  of  the  setting 
of  the  slugs  of  coils  Li  and  Ls. 

When  Cs  is  properly  adjusted,  the  12AU7 
tube  should  be  replaced  in  its  socket  and  coils 
Li  and  Ls  returned  for  maximum  output.  The 
temporary  link  connection  between  Li  and  Li 
should  be  removed  and  the  diode  modulator 
reconnected  into  the  circuit. 

Next,  remove  the  r-f  tubes  and  insert  the 
three  audio  tubes  in  their  respective  sockets. 
Apply  a low  level  audio  signal  of  1225  c.p.s. 
to  the  microphone  jack  Ji  of  the  unit  and  con- 
nect the  horizontal  deflection  terminal  of  the 
oscilloscope  to  a cathode  (pin  3)  of  the 
12AU7  audio  phasing  amplifier,  and  the  ver- 
tical deflection  terminal  to  the  other  cathode 
(pin  8)  after  making  certain  that  the  ’scope 
is  phase-compensated  at  the  test  frequency  of 
1225  c.p.s.  Adjust  audio  balance  control  Rg 
to  produce  a circle  on  the  screen.  Make  this 
test  at  as  low  an  audio  level  as  possible. 

Now,  plug  in  the  r-f  tubes  and  connect  the 
dummy  load  to  the  antenna  receptacle,  Ji. 


574  Low  Power  Transmitters 


THE  RADIO 


Connect  the  vertical  plates  of  the  oscilloscope 
(no  ’scope  amplifier  used)  to  the  terminais 
of  the  dummy  load  so  that  the  80  meter  signal 
of  the  transmitter  may  be  seen  on  the  screen  of 
the  ’scope.  Deliberately  unbalance  one  of  the 
diode  modulators  by  setting  potentiometer  Rs 
off-center.  Adjust  amplifier  output  coil  for 
maximum  pattern  deflection  at  any  convenient 
sweep  speed.  As  the  pattern  grows  when  Ls  is 
resonated  it  may  be  necessary  to  reduce  modu- 
lator unbalance  to  keep  from  overloading  the 
output  stage  or  the  'scope.  Remove  all  audio 
input  to  the  transmitter  by  adjusting  audio 
gain  control  Rs  to  zero,  then  by  successive  ad- 
justments of  the  modulator  balance  controls 
adjust  for  zero  output  as  seen  on  the  oscillo- 
scope. It  will  be  noted  that  as  the  correct  point 
is  approached  the  adjustments  will  become  pre- 
gressively  sharper.  Potentiometers  R«  and  Ro 
should  be  juggled  until  the  indicated  carrier 
is  a minimum.  Slight  capacity  unbalances  with- 
in the  modulator  bridge  may  prevent  a perfect 
minimum  from  being  obtained.  It  may  be  nec- 
essary to  place  a small  10  ggfd  variable  cer- 
amic capacitor  from  either  end  of  potentiome- 
ter Rio  back  to  one  side  of  the  diode  rectifier 
bank  for  best  attenuation.  Good  carrier  at- 
tenuation may  finally  be  achieved  by  juggling 
the  settings  of  Rs,  R»,  and  Rio,  together  with 
the  added  padding  capacitor. 

Next,  apply  a 1225  c.p.s.  audio  signal  and 
advance  gain  control  Rs.  An  r-f  envelope 
should  be  seen  on  the  screen  of  the  oscilloscope. 
The  percentage  of  modulation  (or  ripple) 
noted  on  the  pattern  is  an  indication  of  the 
degree  of  mis-adjustment  of  the  balancing 
controls  of  the  unit.  The  balance  potentiome- 
ters should  now  be  touched  up  to  produce 
minimum  modulation  on  the  observed  carrier. 
Audio  balance  control  Rs  should  also  be  check- 
ed for  minimum  ripple  pattern. 

It  is  a good  idea  to  vary  the  audio  level 
input  to  the  transmitter  to  determine  the  point 
of  overload.  This  can  be  seen  on  he  'scope 
as  the  point  at  which  the  modulation  ripple 
on  the  pattern  increases  sharply  as  Rs  is  ad- 
vanced. Always  make  sure  that  your  audio 
level  is  set  well  below  this  overload  point  when 
adjustments  are  being  made.  It  should  be  pos- 
sible to  produce  a pattern  having  very  little 
ripple  on  the  observed  carrier  wave.  This  com- 
pletes the  adjustments  of  the  transmitter.  Car- 
rier insertion  for  test  purposes  may  be  obtained 
by  reversing  switch  S=  and  adjusting  carrier 
insertion  control  Ri. 

VOX  Circuits  Voice  control  of  the  "send- 
receive”  circuits  of  the 


transmitter  are  desirable  for  serious  sideband 
work,  and  are  doubly  suited  for  mobile  opera- 
tion, since  full  attention  may  be  given  to  road 
problems.  A simple  voice  control  (VOX)  cir- 
cuit suitable  for  use  with  this  transmitter  is 
shown  in  figure  6.  Connection  to  the  speech 
amplifier  is  made  at  point  A of  figure  3.  The 
voice  signal  is  amplified  by  a triode  stage  and 
rectified  by  the  6AL5  diode.  The  resulting 
signal  is  applied  to  the  second  half  of  the 
triode  tube  which  operates  a sensitive  relay 
connected  in  the  plate  circuit.  The  relay  con- 
tacts operate  the  changeover  circuits  of  the 
transmitter.  The  time  delay  required  to  make 
the  transmitter  "hang  on”  between  syllables 
of  speech  is  determined  by  the  R-C  network 
in  the  grid  of  the  relay  control  tube.  Increas- 
ing the  size  of  the  capacitor  will  increase  the 
delay  time  and  vice-versa. 

28-2  A Mobile 
Transistorized 
SS6  Exciter 


Power  consumption  of  radio  equipment  is 
of  paramount  importance  in  mobile  systems- 
Without  the  addition  of  auxiliary  expensive 


Figure  9 

TEN  THOUSAND  MILES  OF  QSO'S 
HAVE  BEEN  MADE  WITH  THIS 
TRANSISTORIZED  FORTY  METER 
SSB  EXCITER 

The  reliabiUty  of  transistorized  sideband  equip- 
ment has  been  proven  with  this  miniature 
phasing-type  exciter.  Overall  size  is  only 
71/4**  X X 71/2".  Antenna,  power,  and  mi- 
crophone connectors  are  on  end  of  chassis 
and  balancing  potentiometers  and  buffer  base 
inductor  (Li)  are  mounted  on  side  of  chassis. 


HANDBOOK 


Transistorized  Exciter  575 


charging  components,  the  usual  automotive 
electrical  system  is  capable  of  a relatively  small 
additional  power  drain  over  and  above  the 
standard  equipment  provided  with  the  car. 
Single  sideband  affords  an  excellent  oppor- 
tunity for  the  mobile  enthusiast  to  obtain 
more  "talk  power”  for  a given  degree  of  pri- 
mary power  consumption.  In  addition,  if  the 
sideband  equipment  can  be  transistorized  the 
power  drain  will  be  reduced  by  a further  im- 
portant amount. 

Described  in  this  section  is  a 40  meter  ex- 
perimental transistorized  sideband  exciter  de- 
livering sufficient  output  to  excite  a tetrode 
linear  amplifier  to  one  kilowatt  peak  input 
power.  The  exciter  operates  directly  from  the 
12  volt  ignition  system  of  the  car,  using  five 
inexpensive  transistors,  a printed  circuit  tran- 
sistor amplifier,  and  a single  vacuum  tube. 
Until  a power  transistor  is  developed  for  r-f 


service,  it  appears  that  the  vacuum  tube  will 
still  be  necessary  as  a buffer  when  high  fre- 
quencies are  involved.  Power  requirements  for 
the  exciter  are  300  volts  at  50  milliamperes, 
and  12.6  volts  at  0.43  amperes. 

A new  subminiature  printed  circuit  ampli- 
fier (Centralab  TA-II)  is  used  for  the  speech 
amplifier.  Completely  sealed  in  epoxy  resin 
for  protection  against  mechanical  abrasion  and 
humidity,  this  four  stage  transistor  amplifier 
provides  a gain  of  over  70  decibels.  The  com- 
plete unit  is  only  114  inches  long  and  less 
than  one  inch  high. 

It  is  strongly  recommended  that  spare  tran- 
sistors be  on  hand  for  substitution  in  the  var- 
ious circuits.  This  will  give  the  builder  confi- 
dence and  assurance  that  each  stage  is  function- 
ing as  it  should.  Many  times  limitations  on  per- 
formance are  imposed  by  certain  transistors 
without  the  operator  being  aware  of  the  con- 


Cl — 2Sp.{ifd.  Variable  ceramic  capacitor.  Certtrafab  823-DZ 
— 35  turns  lt22  enamel,  diam.,  closewound  on  National  Xft-50  form.  Start  with  two  turn  bifitar 
winding  (see  figure  8,  section  28-7  for  bifilar  detail).  Mount  coil  assembly  in  smalt  shield  can. 
i-2,  7-. — Some  as  L,.  Mount  Ls  in  small  shield  can. 

L.^ — Same  as  Lu  Tap  18  turns  from  "cold"  end.  Mount  in  small  shield  can. 

RFC — 2.5  mh.  miniature  r-f  choke.  Millen  J300  - 2500 

Ti — IK  pri.,  60  ohm  sec.  transistor  transformer.  Chicago  UM-111 

T: — 95K  sec.,  ISK  pri.  (use  backwards)  VrC-07 

CHt — 200K  pri.,  1 K sec.  transistor  transformer.  Use  primary  only.  Chicago  UM-112 

X — 7.2  - 7.3  Me  crystal 

Di’Dj. — Matched  lf481  Diodes;  see  text. 


576  Low  Power  Transmitters 


THE  RADIO 


dition.  The  concept  of  transistor  action  is  new 

and  novel  to  many  amateurs;  it  is  not  easy  to 

run  to  the  corner  radio  store  and  test  a transis- 
tor to  make  sure  it  is  operating  properly. 

Exciter  The  complete  schematic  of  the 

Circuitry  transistorized  exciter  is  shown  in 
figure  10.  It  is  derived  from  the 
vacuum  tube  exciter  circuit  of  figure  3,  section 
1,  and  employs  phasing  technique  on  the  fre- 
quency of  operation. 

A 2N112  transistor  is  used  as  a grounded 
emitter  crystal  oscillator  employing  crystals  in 
the  7.2  - 7.3  Me.  range.  The  r-f  output  is 
coupled  through  a tuned  circuit  and  a bifilar 
winding  to  the  crystal  diode  balanced  modu- 
lator. The  output  in  turn  drives  a 2N114  r-f 
amplifier  stage,  operating  in  grounded  emitter 
configuration.  A base  bias  potentiometer  is 
included  in  this  stage  which  permits  parameter 
adjustment  for  maximum  output.  The  collec- 
tor of  the  transistor  is  tapped  at  a point  on  the 
grid  inductor  of  the  linear  amplifier  stage  pro- 
viding maximum  drive  without  exceeding  the 
transistor  ratings. 

A 12BY7A  is  used  as  the  linear  amplifier 
stage.  Neutralization  is  employed  for  greatest 
circuit  stability,  and  VHP  parasitic  oscillations 
are  eliminated  by  the  use  of  a parasitic  sup- 
pressor in  the  plate  circuit. 

The  subminiature  transistorized  audio  am- 
plifier used  in  this  exciter  has  a rising  fre- 
quency characteristic  that  is  well  suited  for 
SSB  work.  However,  the  rising  response  car- 
ries well  into  the  upper  audio  range  wherein 
the  phase  shift  networks  of  the  PS-1  unit 


cease  to  function  properly.  It  is  therefore  nec- 

essary to  incorporate  a simple  low-pass  filter 

that  will  substantially  eliminate  all  audio  fre- 
quencies above  3000  cycles.  Such  a filter  is  in- 
corporated in  the  circuit  between  the  micro- 
phone input  jack  Jz  and  the  transistorized 
audio  amplifier.  The  filter  consists  of  the  pri- 
mary winding  of  a small  audio  transformer 
(CHi)  and  two  250  ggfd  capacitors. 

The  output  signal  from  the  encapsulated 
transistorized  amplifier  is  transformer  coupled 
to  a 2N250  (or  a 2N255)  power  transistor 
which  serves  as  a driver  for  the  phase  shift 
network.  The  2N250  transistor  requires  a 
"heat  sink”  for  the  collector  to  allow  maximum 
collector  dissipation  rating  to  be  achieved.  The 
transistor  may  therefore  be  mounted  on  ano- 
dized aluminum  discs  or  mica  wafers  (sup- 
plied by  the  manufacturer  of  the  transistor ) , 
thus  insulating  the  case  of  the  transistor  (the 
collector  terminal)  from  the  chassis  of  the  ex- 
citer. The  phase  shift  network  driven  by  this 
transistor  is  identical  to  that  described  in  sec- 
tion 28-1  and  illustrated  in  figure  7- 

Two  2N170  NPN  type  transistors  are  base 
driven  by  the  phase  shift  network,  and  serve 
as  emitter  followers,  providing  a low  imped- 
ance driving  source  for  the  balance'd  modulator. 
Sideband  selection  is  accomplished  by  switch 
$2  whose  purpose  is  to  reverse  the  phase  of 
the  audio  signals  applied  to  the  balanced  mod- 
ulator. Carrier  injection  is  accomplished  by 
potentiometer  Rs  and  switch  Si  which  apply 
conducting  bias  to  the  modulator  diodes. 

A voltage  dropping  network  is  employed 


Figure  11 
UNDER-CHASSIS 
VIEW  OF 
TRANSISTORIZED 
SSB  EXCITER 

The  audio  and  r-f  net- 
works are  mounted  on  a 
phenolic  sheet  attached 
to  the  bottom  of  the 
chassis.  The  12BY7A  lin- 
ear amplifier  and  shield 
are  mounted  on  a small 
aluminum  bracket  at  the 
left  of  the  assembly. 
Grid  coil  La  is  beside  the 
tube,  and  plate  coil  Li 
is  behind  the  tube.  Coil 
Li  is  to  the  right  of  La 
with  sideband  switch  St 
behind  it.  The  two  r-f 
balancing  controls  Ri  and 
Ri  are  at  the  extreme 
right  of  the  chassis  panel. 


HANDBOOK 


Transistorized  Exciter  577 


to  reduce  the  12.6  volt  primary  supply  to  9 
volts  for  the  2N170  stage,  and  to  4 volts  and 
2V2  volts  for  the  preamplifier  stage.  The  ex- 
citer is  designed  for  use  with  a primary  supply 
having  the  positive  terminal  of  the  battery 
grounded. 

Exciter  The  exciter  is  built  within 

Construction  an  aluminum  case  measur- 

ing 71/4"  X 4"  X 2V2"  in 
size.  Miniaturization  of  the  audio  phase  shift 
network  was  accomplished  by  measuring  a 
quantity  of  resistors  and  capacitors  on  a bridge. 
Final  selection  was  made  from  those  having 
the  closest  tolerances.  These  hand  picked  com- 
ponents are  mounted  on  a phenolic  terminal 
strip.  This  same  sheet  is  used  to  support  the 
small  r-f  chokes  as  well  as  the  balanced  modu- 
lator components. 

It  is  a good  idea  when  building  equipment 
to  fit  in  a limited  space  to  construct  everything 
in  the  form  of  sub-assemblies.  This  will  permit 
the  components  to  be  mounted  in  the  open, 
facilitating  wiring.  In  this  particular  unit,  the 
audio  components  and  the  r-f  modulator  are 
mounted  on  a phenolic  board  on  one  side  of 
the  box  and  the  r-f  components  are  attached 
to  the  opposite  wall  of  the  box.  The  oscillator 
coil  Li  is  placed  within  an  old  i-f  transformer 
can  as  is  the  linear  amplifier  grid  coil  La.  The 
12BY7A  tube  is  mounted  on  a small  bracket 
at  one  end  of  the  box,  and  has  a tube  shield 
placed  over  it.  The  plate  coil  L.  of  the  amplifier 
stage  is  directly  behind  the  tube  (figure  11). 

Various  voltage  dropping  resistors  are 
mounted  on  a small  terminal  board  at  the  cen- 
ter of  the  chassis,  and  the  audio  transistors 
are  placed  in  the  area  between  the  lower  ter- 
minal board  and  the  rear  wall  of  the  case. 

The  r-f  circuits  are  wired  after  the  com- 
ponent boards  are  mounted  in  place.  Wiring 
is  straightforward  and  simple.  The  two  r-f 
transistors  are  mounted  by  their  leads  to  three 
terminal  phenolic  tie-point  strips.  Be  sure  to 
protect  both  the  transistors  and  the  modulator 
diodes  from  excess  soldering  heat,  as  described 
in  section  28-1.  The  reader  is  referred  to  this 
section  concerning  the  adjustment  and  test 
procedure  for  the  home-made  phase  shift  net- 
work. 

Testing  The  The  transistorized  SSB  exciter 
Exciter  may  be  bench  tested  by  running 

the  transistors  from  a group 
of  IV2  volt  batteries.  It  is  wise  to  conduct 
preliminary  tests  at  an  operating  potential  of 
about  9 volts.  The  voltage  may  be  boosted  to 
the  operating  value  after  tuning  and  align- 
ment adjustments  have  been  completed.  Volt- 


age should  be  first  applied  to  the  transistor 
oscillator.  R-f  output  may  be  measured  at  the 
terminals  of  the  bifilar  coil  with  the  aid  of  a 
vacuum  tube  voltmeter.  The  slug  of  coil  Li 
is  adjusted  for  maximum  reading.  Under  diode 
load,  about  4 volts  r.m.s.  may  be  measured 
from  either  side  of  the  winding  to  ground. 
The  bifilar  winding  is  relatively  critical.  Too 
many  turns  will  load  the  oscillator  to  a point 
of  instability,  and  too  few  turns  deliver  in- 
sufficient drive  to  the  modulator.  Two  turns, 
loosely  coupled  to  the  primary  winding,  seem 
to  be  about  correct.  When  the  bridge  is  un- 
balanced, approximately  0.2  volt  r.m.s.  may  be 
measured  from  either  arm  of  output  potentio- 
meter Rs  to  ground.  This  voltage  is  stepped 
up  by  means  of  a pi-network  circuit  so  that 
maximum  excitation  voltage  is  applied  to  the 
base  of  the  2N114  r-f  transistor. 

Preliminary  adjustments,  in  general,  follow 
those  described  for  the  vacuum  tube  exciter 
described  in  section  28-1.  Potentiometers  Ri, 
R2,  and  R.1  are  adjusted  for  minimum  carrier 
output  in  the  collector  circuit  of  the  2N114. 
It  may  be  necessary  to  add  a 10  ggfd  variable 
ceramic  capacitor  from  one  side  of  the  modu- 
lator bridge  to  ground  to  achieve  maximum 
carrier  suppression.  The  final  adjustment  is  to 
apply  an  audio  signal  and  touch  up  the  audio 
phasing  potentiometer  R»  for  maximum  re- 
jection of  the  unwanted  sideband. 

When  these  preliminary  adjustments  have 
been  made,  the  12BY7A  tube  may  be  inserted 
in  the  socket,  and  the  shield  slipped  over  the 
tube.  Grid  inductor  La  is  resonated  for  maxi- 
mum grid  excitation  with  the  aid  of  grid  dip 
oscillator,  and  the  plate  inductor  Li  can  be 
tuned  in  the  same  manner.  A dummy  load 
consisting  of  two  6.3  volt,  150  ma.  flashlamp 
bulbs  (brown  bead)  connected  in  parallel  is 
attached  to  output  jack  Ji.  Plate  voltage  and 
excitation  are  now  applied  to  the  linear  ampli- 
fier stage  and  La  and  La  adjusted  for  maximum 
output.  When  audio  excitation  is  removed 
from  the  exciter  the  r-f  amplifier  will  prob- 
ably break  into  oscillation  and  the  antenna 
bulbs  will  continue  to  glow.  Neutralizing  ca- 
pacitor Cl  should  now  be  adjusted  with  a 
fibre  screwdriver  until  oscillation  stops.  After 
the  exciter  is  operating  properly,  full  battery 
voltage  may  be  applied  to  the  transistor  stages 
with  a consequent  increase  in  grid  drive  and 
power  output  of  the  linear  amplifier.  Approx- 
imately 2 volts  r.m.s.  grid  drive  may  be  meas- 
ured with  the  vacuum  tube  voltmeter  at  the 
grid  pin  of  the  12BY7A  linear  amplifier  stage 
under  conditions  of  maximum  excitation. 


578  Low  Power  Transmitters 


THE  RADIO 


28-3  A VH F T ransceiyer 
of  Advanced  Design 

Modern  tubes  and  circuit  techniques  permit 
sensitivity  and  stability  levels  to  be  achieved 
in  the  VHF  region  that  were  possible  only  at 
the  lower  frequencies  a few  years  ago.  The 
transceiver  described  in  this  section  makes  full 
use  of  these  advanced  circuits  to  produce  su- 
perb results  in  the  amateur  50  Me.  and  144 
Me.  bands. 

The  transceiver  (pictured  in  figure  12) 
comprises  a complete  VHF  station  and  power 
supply  in  one  small,  streamlined  cabinet.  The 
transmitter  section  is  crystal  controlled  on  any 
one  of  eleven  frequencies  chosen  at  random 
in  the  two  bands,  and  runs  eighteen  watts  in- 
put fully  modulated.  The  receiver  section  em- 
ploys cascode-type  r-f  amplifiers  resulting  in  a 
noise  figure  better  than  5 decibels  on  both 
bands.  Optimum  communications  selectivity  is 
obtained  by  the  use  of  a two-section  crystal 
lattice  filter  in  the  intermediate  frequency  am- 
plifier. Maximum  frequency  stability  is  achiev- 
ed by  the  use  of  crystal  controlled  VHF  con- 
version oscillators  and  a highly  stable  low  fre- 
quency tunable  oscillator.  A signal  strength 
meter  is  incorporated  in  the  receiver,  and  a 
tuning  meter  is  used  to  obtain  optimum  trans- 
mitter adjustment.  The  power  supply  operates 
from  115  volts  a.c.,  and  either  6-  or  12-volts 
d.c. 

Styling  for  the  transceiver  permits  its  use 
either  as  a fixed,  home  station  or  as  a mobile 
unit.  The  major  tuning  controls  are  placed  at 
the  left  of  the  panel  to  permit  operation  while 
the  automobile  is  being  driven.  A push-to-talk 
system  is  employed  for  rapid  break-in  opera- 
tion, and  band  changing  takes  place  in  a few 


seconds.  Twenty  tubes  including  a voltage  reg- 
ulator are  employed  in  the  unit,  and  the  total 
primary  power  consumption  is  approximately 
105  watts. 

Circuit'  This  high  performance  trans- 

Descripiion  ceiver  is  a complex  device,  and 
the  wiring  diagram  has  been 
broken  down  into  several  schematic  drawings 
(figures  14,  15,  17,  18,  and  19).  A block  dia- 
gram of  the  complete  circuit  is  given  in  figure 

13. 

The  Receiver  Section.  The  receiver  portion 
of  the  unit  employs  thirteen  tubes  in  a double 
conversion  superheterodyne  circuit.  Separate  r-f 
"front  ends”  are  employed  for  each  band.  Each 
"front  end”  employs  a 6BZ7  cascode  t-f  am- 
plifier and  a 6J6  converter/oscillator  tube.  The 
desired  section  is  activated  by  energizing  the 
filaments  and  connecting  the  output  circuit 
of  the  mixer  stage  to  the  first  i-f  amplifier. 
Switch  Si  accomplishes  this  action.  The  144 
Me.  cascode  amplifier  is  neutralized  for  opti- 
mum signal-to-noise  ratio,  while  the  neutraliz- 
ing circuit  is  omitted  from  the  50  Me.  ampli- 
fier since  a satisfactory  noise  figure  is  obtained 
without  the  extra  coil.  Overtone  crystals  oper- 
ating in  the  VHF  region  are  used  in  a regen- 
erative oscillator  circuit  to  provide  mixing 
voltage  for  the  6J6  triode  converter.  The  com- 
plete schematic  of  the  r-f  section  is  given  in 
figure  14. 

The  tunable  intermediate  frequency  ampli- 
fier covers  the  range  of  30  - 34  Me.  "Low  side” 
oscillator  operation  is  used  for  two  meter  oper- 
ation with  the  conversion  crystal  at  114  Me. 
"High  side”  operation  is  used  for  six  meter 
service  with  the  crystal  at  84  Me.  The  four 
megacycle  i-f  tuning  range  is  covered  by  a 


Figure  1 2 

VHF  Transceiver  for  Six  and 


Two  Meters  is  Designed  for 
Fixed  or  Mobile  Operation 
This  transmitter-receiver  pro- 
vides the  highest  order  of  re- 
liable communication  in  the  two 
most  popular  VHF  bands.  Re- 
ceiver tuning  controis  are 
placed  at  the  left  of  the  unit 
to  permit  ease  of  tuning  in 
mobile  installation.  Double  con- 
version receiver  and  crystal  lat- 
tice i-f  filter  provide  maximum 
stability  and  selectivity  for  re- 
ception. Crystal  controlled  trans- 
mitter delivers  a ten  watt  car- 
rier. Band  change  takes  only  a 
few  seconds.  Left  hand  meter  is 
signal  strength  indicator  for  re- 
ceiver, and  meter  at  right  is 
multi-purpose  millommeter  for 
transmitter.  The  transmitter  tun- 
ing controls  are  at  right.  Perfo- 
rated metal  cabinet  insures 
maximum  ventilation  for  con- 
tinuous operation. 


HANDBOOK 


Advanced  Transceiver  579 


TV-type  rotary  induction  tuner,  padded  to  the 
correct  operating  range.  The  use  of  this  type 
of  tuned  circuit  provides  exceptional  stability 
combined  with  small  size  and  has  proven  to 
be  a very  satisfactory  choice.  A 6BJ6  remote 
cutoff  pentode  is  used  as  the  first  i-f  amplifier 
stage  (Vs)  and  a 6BE6  is  used  as  the  second 
mixer  (Vs).  The  tunable  oscillator  stage  op- 
erates 2009  kilocycles  lower  than  the  variable 
intermediate  frequency  and  employs  a single 
section  of  a 6j6  tube  (Vi).  A small  inductance 
(Lii)  is  placed  in  series  with  the  rotary  coil 
of  the  oscillator  section  to  ensure  proper  track- 
ing over  the  4 Me.  tuning  range. 

The  greater  portion  of  the  receiver  gain  and 
selectivity  is  achieved  in  the  second  inter- 
mediate frquency  amplifier  which  operates  at 
a center  frequency  of  2009  kc.  A double  lat- 
tice crystal  filter  is  employed  to  ensure  maxi- 
mum selectivity  at  this  frequency.  The  filter 
is  made  in  two  sections  and  has  a single  6BJ6 
amplifier  tube  (Vs)  between  the  sections  to 
compensate  for  the  insertion  loss  of  the  filter. 

The  first  section  of  the  filter  is  placed  im- 
mediately after  the  6BE6  converter  tube  to 
achieve  maximum  i-f  amplifier  protection  from 
strong  signals  adjacent  to  the  desired  one.  Two 


crystals  are  used,  one  at  a frequency  of  2007 
kc.  and  the  other  at  a frequency  of  2011  kc., 
providing  a "flat  top”  to  the  passband  of  about 
4 kilocycles.  The  crystals  used  in  this  filter  are 
specially  cut  for  filter  work,  and  the  substitu- 
tion of  ordinary  "transmitter  type”  crystals  will 
degrade  the  performance  of  the  filter.  The  units 
specified  in  figure  15,  however,  are  not  ex- 
pensive. The  i-f  transformers  are  standard  1500 
kc.  units  with  the  mica  tuning  capacitors 
changed  as  shown  in  figure  15  to  permit  oper- 
ation at  2 Me.  The  lattice  filter  may  be  re- 
moved from  the  circuit  by  means  of  the  "sharp- 
broad”  switch.  Si.  All  four  crystals  are  shorted 
out  in  the  "broad”  position,  and  one  crystal  in 
each  filter  section  is  disconnected  from  the  as- 
sociated i-f  transformer.  In  addition,  another 
section  of  the  switch  (SiC)  drops  the  gain 
of  the  amplifier  tube  (Vs)  to  compensate  for 
the  rise  in  stage  gain  that  occurs  when  the 
crystals  are  removed  from  the  circuit.  Two 
additional  stages  of  i-f  amplification  (Vs,  Vm) 
follow  the  crystal  filter,  providing  an  abund- 
ance of  gain  for  proper  operation  of  the  a-v-c 
and  audio  squelch  circuits. 

The  detector  and  associated  audio  compon- 
ents of  the  receiver  are  shown  in  figure  17.  A 


Vi  Va 


Twenty  tubes  plus  selenium  rectifier  power  supply  provide  de-luxe  operation  for  the  advanced  VHF 
enthusiast.  Power  supply  may  be  operated  from  either  automobile  electrical  system  or  115  volts,  a.c. 


580  Low  Power  Transmitters 


THE  RADIO 


6AL5  (Vn)  is  employed  as  detector,  a-v-c, 
and  a-n-1  tube.  The  noise  limiter  is  of  the  low 
distortion  type  and  is  left  in  the  circuit  at  all 
times.  One  section  of  the  6AL5  serves  as  the 
a-v-c  rectifier,  and  the  resulting  voltage  is  ap- 
plied to  four  i-f  tubes  for  smooth  a-v-c  action 
on  strong  signals.  The  last  i-f  stage  (Vio)  is 
left  off  the  a-v-c  line,  and  the  S-meter  is  con- 
nected between  the  plate  supply  return  of  this 
tube  and  an  adjacent  controlled  tube.  A change 
in  the  plate  current  of  the  controlled  tube 
(V»)  upsets  the  balance  of  the  bridge  circuit, 
and  a meter  reading  is  produced  that  is  pro- 
portional to  the  a-v-c  voltage.  This  voltage  is 
also  applied  to  one  section  of  a double  triode 
audio  tube  (V12)  which  functions  as  a squelch 
tube.  This  circuit  renders  the  audio  section  of 
the  receiver  inoperative  until  the  receiver  a-v-c 
circuit  is  activated,  thus  eliminating  a large 
portion  of  the  background  noise  commonly  as- 
sociated with  VHF  reception.  The  squelch  cir- 
cuit may  be  disabled  by  switch  Sa.  Two  audio 


stages  (V12  and  Vm)  deliver  sufficient  audio 
power  to  fully  drive  a loudspeaker  and  to  over- 
come the  motor  and  wind  noises  of  a car  in 
motion.  An  auxiliary  jack  (J3)  provides  ear- 
phone reception  if  desired.  The  receiver  is  in- 
activated during  transmission  periods  by  relay 
contacts  RY2B  which  remove  the  high  voltage 
from  the  plates  and  screens  of  the  receiver 
tubes. 

The  Transmitter  Section.  Shown  in  figure  17 
is  the  transmitter  section  of  the  transceiver. 
Six  tubes  are  used,  three  for  the  audio  portion 
of  the  unit  and  three  for  the  r-f  portion. 
Crystals  in  the  24-25  Me.  region  are  employ- 
ed for  both  6-  and  2-meter  operation.  Any  one 
of  eleven  crystals  may  be  selected  by  switch 
Si.  The  oscillator  is  a 6CL6  (Vis)  in  a "hot- 
cathode”  doubler  circuit,  delivering  energy  in 
the  48  - 50  Me.  region.  A split-stator  tuning 
capacitor  is  employed  in  the  plate  circuit  to 
provide  a balanced  output  configuration.  The 
oscillator  stage  may  be  turned  on  during  re- 


Vi  va 

6BZ7A  M4  MC.  section  6J6 


Figure  14 


SCHEMATIC/  R-F  PORTION  OF  RECEIVER  SECTION 

C — .001  iifd.  d/sc  ceramic  capacitor.  Centralab  DD-102 

C],  Cl — 20  /xnfd.  Johnson  20M11 

Ji,  Ji — BNC-type  coaxial  receptacle 

Li-Ls — See  figure  21  for  coil  information 

S7A,  D — Pari  of  57  (Centralab  wafer  PA-9)  See  figure  18. 

RYiA,  RYiA — See  figure  19 

Xi — 114  Me.  overtone  crystal.  Precision  Crystal  Lab.,  Santa  Monica,  Calif. 
Xs — 84  Me.  crystal  (see  X;) 


Figure  15 

SCHEMATIC,  l-F  PORTION  OF  RECEIVER  SECTION 

Cs,  Ci — 30  ceramic  trimmer.  Centraiab  827-C 

Cs — 12  fi/ifd.  ceramic  trimmer.  Centraiab  NPO 
L:o-Lii — See  figure  21  for  coil  information 
Ml — 0>l  d.c.  milliammeter.  De-fur  #112 
RYiB,  RYiC—^ee  figure  19 

SiA-SiE — 2 wafers,  6 pole,  2 position  each  wafer.  (Two  Centraiab  PA-9  wafers  with  PA-301  index  assembly) 
Ti-Ts — Modified  1500  kc.  slug-tuned  interstage  i-f  transformer.  (J.W.  Miller  12-W1),  Remove  internal 
resonating  capacitors.  Replace  plate  capacitor  with  50  /ififd.  mica,  and  grid  capacitor  with  two 
100  fi/xfd.  mica  capacitors  connected  in  series.  Ground  common  tap  of  capacitors.  Transformer  now 
tunes  to  2009  kc. 

Ti — 1500  kc.  slug-tuned  i-f  output  transformer.  (J.  W.  Miller  12-W2}.  Modify  as  above. 

X3-X1. — Lattice  crystal  set.  2007  kc.  and  2011  kc.  Precision  Crystal  Labs.,  Santa  Monica,  Calif. 

X5-X5*— Same  as  X3-X1,. 

RFC— 39  nh.  J.  W.  Miller  4628 


Advanced  Transceiver  581 


582  Low  Power  Transmitters 


THE  RADIO 


ception  by  depressing  the  "zero”  switch  Se, 
which  applies  plate  voltage  to  the  oscillator 
by  way  of  the  receiver  circuits. 

The  balanced  output  circuit  of  the  oscillator 
stage  is  inductively  coupled  to  a push-pull 
miniature  beam  amplifier  stage  employing  a 
6360  (Vic).  The  grid  circuit  of  this  stage  is 
broadly  resonant  to  the  50  Me.  region  by  vir- 
tue of  the  interelectrode  capacitances  of  the 
tube.  For  six  meter  operation,  the  6360  func- 
tions as  a class-C  amplifier  with  the  plate  cir- 
cuit (Lis-C?)  tuned  to  50  Me.  Bias  and  screen 
voltages  are  set  for  optimum  operating  values 
by  segments  of  switch  Si.  In  addition,  the  plate 
return  of  the  stage  is  routed  through  the  sec- 
ondary of  modulation  transformer  Tz.  The  r-f 
output  circuit  is  inductively  coupled  to  the  50 
Me.  antenna  coaxial  receptacle  through  relay 
contacts  RY2A.  , 

When  switch  Si  is  moved  to  the  144  Me. 
position,  the  aforementioned  6360  stage  is  con- 
verted into  a high  efficiency  triplet  stage,  pro- 
viding output  in  the  140-  150  Me.  region. 
Switch  segment  SaD  raises  the  operating  bias 
on  the  stage,  segment  SaE  drops  the  screen 
voltage,  and  segments  SaA-B  select  the  proper 
plate  tank  coil,  Lu. 

A separate  amplifier  stage  employing  an- 
other 6360  dual  beam  power  tube  (Via)  is 
used  for  2-meter  operation.  The  tube  functions 
as  a class-C  amplifier  and  is  driven  by  the 
push-pull  triplet  stage  (Vaa).  The  grid  circuit 
of  the  144  Me.  amplifier  is  inductively  coupled 


to  the  triplet,  and  the  plate  circuit  (G-Lis)  is 
inductively  coupled  to  the  output  circuit.  A 
separate  antenna  receptable  (Ji)  is  used  for 
144  Me.  operation.  Plate  voltage  is  applied  to 
this  tube  by  switch  section  SaF. 

The  use  of  inductive  coupling  between  the 
three  stages  of  the  transmitter  reduces  spurious 
emissions  to  a minimum.  Channel  2 television 
interference,  so  common  to  six  meter  trans- 
mitters is  absent  when  this  transmitter  is  used, 
thanks  to  the  combination  of  interstage  induc- 
tive coupling  and  the  employment  of  25  Me. 
crystals.  Power  input  to  the  modulated  stage 
on  both  bands  is  close  to  18  watts,  permitting 
a carrier  power  of  over  1 0 watts.  , 

Transmitter  adjustment  is  greatly  simplified 
by  the  use  of  r-f  volt-meters  connected  to  the 
output  circuits  of  each  amplifier  stage.  Small 
germanium  diodes  (Di  and  D2)  are  used  to 
rectify  a minute  portion  of  the  output  voltage 
which  is  read  on  the  multi -meter  Mi  of  the 
transmitter.  All  tuning  adjustments  of  the 
transmitter  are  conducted  so  as  to  enhance  the 
reading  of  the  r-f  voltmeter  under  a given  set 
of  antenna  conditions. 

The  transmitter  is  plate  modulated  by  two 
6AQ5  tubes  (Vi»,  V20)  operating  class  ABi. 
Plate  voltage  may  be  removed  from  these  tubes 
and  from  the  modulated  amplifier  by  the  "tune- 
operate”  switch  Sj.  A single  12 AX  7 serves  as 
a two  stage  resistance  coupled  audio  amplifier, 
having  sufficient  gain  for  the  use  of  a crystal 
or  dynamic  microphone- 


IHJWWa 
mu  i 


Figure  16 

REAR  VIEW  OF 
VHF  TRANSCEIVER 

At  the  left  rear  of  the  chassis  is 
the  dual  purpose  power  supply 
with  the  vibrator  and  filter  ca- 
pacitors mounted  in  front  of  the 
power  transformer.  At  the  front 
of  the  chassis  are  the  two  6360 
tubes  and  the  tuning  capacitors  of 
the  transmitter.  Note  that  the 
stages  ore  isolated  by  shields 
traced  between  the  tuning  capaci- 
tors and  tubes. 

The  modulator  and  power  supply 
filter  choke  occupy  the  center  por-  / 
tion  of  the  chassis,  with  the  re- 
ceiver i-f  section  to  the  right.  In 
front  of  the  i-f  strip  are  the  trans- 
mitter crystals  and  the  lattice 
crystals.  At  the  right  of  the  chas- 
sis are  located  the  '*TV-type” 
variable  inductor  tuner  and  the 
three  tunable  i-f  stage  tubes.  At 
the  rear  of  the  chassis,  directly 
behind  the  tuner  are  the  two  VHF 
conversion  oscillators.  The  micro- 
phone receptacle  and  audio  gain 
control  are  on  the  rear  lip  of  the 
chassis,  directly  below  the  cascode 
stages.  Cabinet  and  chassis  cus- 
tom made  by  California  Chassis 
Co.,  Lynwood,  Calif. 


i 


M2*  0-10  MA. 
M3.M4S0-100MA. 


f ^ 

“■I 

120 
c _ 

iC?  Li3oI 
J»  0) 

. TO  RY2A 
TIC.  15 


f H( 


120 
c _ 

C9 

rY  RFCE 


TO  RY2B 
Fig. 15 


,«io Tin 


yS  RFC  2 
Sae 


(0)9  <2)  iSOK 

?T  Sac,^ 


Figure  17 

SCHEMATIC,  TRANSMITTER  AND  AUDIO  SECTION  OF  TRANSCEIVER 

C5,  C-,  Cj — 75/15  /i/uW.  Hammarlund  HfD-ISX  SjA-SsF — Some  os  Si,  figure  75.  Space  decks  so  coi/s  Li.  and  Lu  fit  between 

Cs,  Citr— 30  fi^fd.  Johnson  2SJ12  S« — SPOT  "push-boffon"  switch 

Di,  Di — Germanium  diode  1N81  5s — 2 pole,  3 position  non-shorting  switch.  Mallory  3223-G 

Lii-Ljs — See  figure  21  for  coil  information  RYt,  RYt — ^ee  figure  19 

RFC- 1 — 7.0  fih.  Ohmite  Z-50  Mf — O-I  d.c.  milliammeter.  De-jur  if  112 

RFCi — 1.8  /iff-  Ohmite  1-144  T2 — 25  watt  modulation  transformer,  10K  pri.,  5 K sec.  Stancor  A-384S 

RFC3 — 0.75  /j.h.  J.  W.  Miller  4644  Tj — Interstage  transformer,  1:3.  Stancor  A-S3C 

Xj— 24-26  Me.  overtone  crystal.  Precision  Crystal  Co.,  Santa  Monica,  Calif. 


fit  between  the  decks. 


Advanced  Transceiver  583 


584  Low  Power  Transmitters 


THE  RADIO 


The  Power  Supply  Section,  The  complete 
transceiver  requires  a plate  supply  of  250  volts 
at  150  milliamperes,  and  either  6.3  volts  at 
7.8  amperes  or  12.6  volts  at  3.9  amperes.  These 
voltages  may  be  obtained  from  the  power  sup- 
ply whose  schematic  is  given  in  figure  18.  The 
high  voltage  portion  of  the  supply  employs  a 
voltage  doubler  selenium-type  rectifier  system 
operating  from  a 117  volt  half-wave  winding 


of  the  power  transformer.  This  transformer  has 
two  auxiliary  windings.  One  winding  may  be 
used  for  115  volt,  50-60  cycle  operation,  and 
the  other  one  is  intended  for  vibrator  opera- 
ation  from  either  a 6-  or  12-volt  d c.  primary 
source.  The  choice  of  a.c.  or  d.c.  operation  may 
be  made  by  inserting  the  correct  terminal  plug 
in  receptacle  SOi,  mounted  on  the  rear  of  the 
transceiver  chassis- 


115V,  A-C 

PL2 


HANDBOOK 


Advanced  Transceiver  585 


The  power  transformer  is  made  from  a TV- 
replacement  type,  with  a new  vibrator  winding 
replacing  the  usual  6.3  volt  filament  windings. 
Information  covering  the  conversion  of  the 
transformer  will  be  found  later  in  this  section. 
The  special  winding  may  be  arranged  for  eith- 
er 6-volt  d.c.  or  12-  volt  d.c.  operation,  but 
not  both.  Thus,  the  choice  of  d.c.  supply  volt- 
age must  be  determined  before  the  transformer 
is  modified. 

The  filaments  of  the  transceiver  are  ar- 
ranged in  a series  parallel  circuit  so  that  they 
may  be  operated  from  either  6-  or  12-volts 
d.c.,  or  6-3  volts  a.c.  with  no  change  in  cir- 
cuitry. The  appropriate  filament  connections  to 
the  power  supply  are  automarically  made  when 
the  proper  terminal  plug  is  inserted  in  SOi.  A 
simplified  drawing  of  the  primary  power  wir- 
ing of  the  transceiver  is  shown  in  figure  18. 

Relay  and  Switching  Circuits.  The  control 
circuits  of  the  de-luxe  transceiver  are  shown  in 
figure  19-  Two  relays  control  change-over  op- 
eration. The  relay  coils  are  of  the  d.c.  type, 
and  are  connected  in  series  with  the  push-to- 
talk  button  of  the  microphone  and  the  filament 
voltage  supply.  Closing  the  microphone  switch 
grounds  one  end  of  the  relay  coil  circuit  and 
actuates  both  relays.  Relay  RYi  controls  the 
144  Me.  antenna  changeover,  and  switches  the 
S-meter  to  the  r-f  output  meter  circuit  of  the 
transmitter  section.  Relay  RYi  conrols  the  50 
Me.  antenna  changeover,  and  switches  the  B- 
plus  from  the  receiver  to  the  transmitter. 

Meter  M2  serves  as  a multi-meter  for  the 
transmitter  portion  ( figure  17).  The  basic 
meter  has  a 0-10  d.c.  millammeter  movement 
and  is  used  as-is  for  measuring  the  grid  current 
of  V16  and  Vn.  Meter  switching  is  accomplished 
by  means  of  Ss.  The  third  position  of  the  switch 
connects  the  meter  across  the  shunt  in  the  plate 
circuit  of  the  modulated  amplifier  stage.  The 
resistance  of  the  shunt  is  chosen  so  that  full 
scale  meter  deflection  is  100  milliamperes. 

Transceiver  Layour  The  transceiver  is  housed 
and  Assembly  in  a special  cabinet  mea- 

suring 14"  long,  10" 
deep,  and  SVj"  high.  The  cabinet  is  formed 
from  "cane  metal”  stock  and  has  a removable 
front  panel.  The  back,  which  is  also  made  of 
solid  material  is  welded  to  the  wrap-around 
cane-metal  cover.  The  chassis  of  the  unit  is 
made  of  sheet  steel  which  is  copper  plated 
after  all  major  holes  have  been  drilled.  Chas- 
sis height  is  2 inches.  Placement  of  the  major 
components  may  be  seen  in  figure  16.  The 
receiver  section  occupies  the  right-hand  por- 


RELAY  SWITCHING  CIRCUIT 
(.SHOWN  UNENERGfZED) 


R Y1 

A ^ 


2 X R-F  COIL 
TO  J 1 

XMTR  ANT.  COIL 


r— o-»  TO  V9 

TO  +S-MeTER 

A o-»  TO  V10 

TO  —S-METER 
TO  R-F  DIODES 


6X  R-F  COIL 
TO  J 2 

XMTR  )S.NT.  COIL 

TO  RECEIVER 
TO  B + 

TO  XMTR 


TO  CONTROL  CIRCUIT 
IN  MICROPHONE 


TO  FILAMENT 
VOLTAGE  SUPPLY 


TRANSCEIVER  CONTROLS 


SWITCHES 


51- SELECTIVITY 

52- XMTR  CRYSTAL 
SS-BANDSWITCH,  XMTR 

54- FILAMENT 

55- TUNE-OPERATE 
Ss- "lERO" TUNE 

S7  - BANDSWITCM,  RECEIVER 
Ss  - METER  SELECTOR,  XMTR 
S9-SQUELCH 

SlO-POWBR  (.BACH  OF  CHASSIS) 


POTENTIOMETERS 

Ri-receiver  volume 

R2-XMTR  AUDIO  GAIN 

Rs-S-meter  ADJ . 


Figure  19 

RELAY  CONTROL  WIRING  AND 
LIST  OF  TRANSCEIVER  CONTROLS 

Ki,,  KYr—3  pole,  DT.  Advance  MF3C/6VD. 
Hook  colls  in  parallel  for  6 volt  operation. 


tion  of  the  chassis,  with  the  power  supply  to 
the  tear  left,  and  the  transmitter  at  the  upper 
left  of  the  photograph.  The  variable  TV-type 
tuner  is  panel  driven  by  the  main  tuning  dial 
through  a simple  speed  reduction  gear  f Crowe). 
To  the  side  of  the  tuner  are  tubes  Vs,  Ve,  and 
V,.  To  the  rear  of  the  tuner  are  the  two  separ- 
ate "front  end”  sections  of  the  receiver.  The  two 
cascode  stages  ate  in  the  corner,  with  the  6J6 
mixer/oscillator  tubes  to  the  left.  The  various 
coil  slugs  and  tuning  capacitors  are  mounted 
to  the  chassis  deck  and  are  adjustable  from  the 
top  of  the  unit.  Near  the  center  of  the  chassis 
and  running  from  the  front  to  the  back  is  the 
2009  kc.  i-f  strip.  The  four  crystals  of  the  i-f 
filter  are  seen  near  the  front  of  the  chassis  di- 
rectly behind  i-f  transformer  Ti.  To  the  left 
of  the  i-f  strip  arranged  in  a line  are  the  volt- 
age regulator  tube  Vn,  the  first  lOgfd.  filter 
capacitor  for  the  speech  amplifier,  the  speech 
amplifier  tube  Vm,  the  second  lOgfd.  filter 
capacitor,  and  the  receiver  audio  amplifier/ 
squelch  tube  Vu.  To  the  left  of  these  com- 
ponents are  the  two  6AQ5  modulator  tubes 
and  transformers  T2  and  Ts.  In  the  far  left 


586  Low  Power  Transmitters 


THE  RADIO 


corner  of  the  chassis  are  placed  the  power  trans- 
former Ti,  the  vibrator  VIBi,  and  the  two 
power  supply  filter  capacitors.  Note  that  the 
case  of  one  of  the  capacitors  is  "hot”  to  ground 
and  the  unit  is  mounted  on  a fibre  base  ring. 

The  transmitter  components  occupy  the  up- 
per left  chassis  area  of  the  transceiver.  The 
6CL6  oscillator  stage  is  at  the  center  of  the 
chassis  with  tuning  capacitor  C«  to  the  left.  A 
small  aluminum  shield  isolates  the  oscillator 
components  from  the  6360  buffer/ amplifier 
tube.  A second  shield  plate  is  placed  between 
the  second  stage  tuning  capacitor  Ci  and  the 
6360  144  Me.  amplifier. 

Space  is  provided  on  the  chassis  to  mount 
five  miniature  crystal  holders  for  the  trans- 
mitter oscillator  circuit.  A small  aluminum 
clamp  holds  five  additional  crystal  sockets  to 
the  rear  of  crystal  selector  switch  Ss,  as  seen 
in  the  under-chassis  photograph  of  figure  20. 

The  power  supply  area  beneath  the  chassis 
is  separated  from  the  remainder  of  the  cir- 
cuitry by  a 2 inch  high  copper  shield  that  en- 
closes the  selenium  rectifiers  and  other  small 
supply  components.  Leads  passing  out  of  this 
compartment  go  through  chassis  mounting  type 
miniature  feedthrough  capacitors  mounted  in 
holes  in  the  shield  {Centralab  FT-1000) . 

The  "broad-sharp”  switch  Si  may  be  seen  in 
the  upper  left  area  of  the  chassis.  Each  segment 
of  the  switch  is  placed  over  a group  of  i-f 
crystal  sockets  so  that  the  leads  from  the  crys- 
tals to  the  switch  contacts  are  extremely  short. 
To  the  left  of  this  switch  is  the  S-meter  "zero” 
potentiometer  Rs.  The  receiver  bandswitch  Si 
is  located  adjacent  to  the  cascode  amplifier 


stages  at  the  rear  of  the  chassis  and  is  actuated 
from  the  panel  by  an  extension  shaft. 

The  r-f  section  of  the  transmitter  is  visible 
in  the  upper  right  area  of  the  chassis,  running 
parallel  to  the  front  panel.  A small  brass 
shield  is  passed  across  the  center  of  each  of  the 
6360  tube  sockets.  The  shield  passes  between 
pins  1 and  9,  and  3 and  4.  Transmitter  band- 
change  switch  Sa  is  located  in  the  center  com- 
partment and  the  adjoining  shields  prevent 
coupling  between  the  coils  of  this  circuit  and 
either  the  144  Me.  amplifier  coil  or  the  50 
Me.  oscillator  coil. 

Transceiver  The  transceiver  is  a complex 

Wiring  piece  of  equipment  and  should 

be  wired  with  care.  Space  is  at 
a premium  and  wiring  errors  are  difficult  to 
find  when  working  in  a small  area.  It  is  there- 
fore suggested  that  the  unit  be  wired  in  sec- 
tions. The  receiver  should  be  wired  first,  and 
placed  in  operation  with  an  auxiliary  power 
supply.  The  transmitter  may  be  wired  next 
and  tested  for  operation  with  the  receiver. 
Finally,  the  power  transformer  should  be  con- 
verted and  the  power  supply  can  be  wired  and 
tested  within  the  transceiver.  The  final  check 
is  to  make  sure  that  all  sections  operate  proper- 
ly as  a complete  system. 

Wiring  the  Receiver  Section.  The  receiver 
section  should  be  wired  first.  It  is  imperative 
that  all  r-f  wiring  and  ground  leads  be  short 
and  direct.  As  many  components  as  possible 
are  mounted  between  the  pins  of  the  tube 
sockets  and  the  socket  ground  lugs.  Miniature 
disc  ceramic  capacitors  are  used  for  bypass  pur- 


Figure  20 

UNDER-CHASSIS  VIEW  OF 
TRANSCEIVER 

General  layout  of  the  components  below  the 
chassis  may  be  seen  in  this  view.  At  the  lower 
right  are  the  power  supply  components,  sep- 
arated from  the  rest  of  the  circuitry  by  a 
copper  shield.  RY 1,  RYr,  and  the  audio  driver 
transformer  are  just  above  this  shield.  Across 
the  front  of  the  chassis  is  the  transmitter 
section  with  the  inter-stage  shields  mounted 
across  the  tube  sockets.  At  the  center  is  the 
crystal  switch  with  five  crystals  mounted  to 
the  wafer  of  the  switch.  Directly  behind  this 
switch  is  the  i-f  section  of  the  receiver  and 
the  audio  output  transformer.  In  the  lower 
left  corner  are  the  cascode  r-f  stages  and  the 
bandswitch  for  the  receiver  which  is  mounted 
on  a small  bracket  and  panel  driven  with 
an  extension  shaft.  In  front  of  this  switch 
are  the  components  of  the  tuned  first  con- 
version stage.  The  trimmers  used  with  the  TV 
"inductuner"  are  mounted  on  the  tuner  ter- 
minals. 

Sections  of  '/4-inch  coaxial  cables  connect  the 
various  antenna  circuits  to  the  change-over 
relays  and  the  coaxial  antenna  receptacles 
on  the  rear  of  the  chassis. 


HANDBOOK 


Advanced  Transceiver  587 


FIGURE  21 

COIL  TABLE  FOR  2 AND  6 METER  TRANSCEIVER 


RECEIVER  SECTION,  FIGURES  14  AND  15 

L1  - 6T.  #Z0  E.,  3/6"  DIA.,  1/2*  LONG.  TAP  2 1/2  T. 

FROM  GROUND  END. 

L2-  1 2 T.  # 20  E,  3/16*  DIA.  7/16*  LONG.  MOUNT  BETWEEN 
PIN  3 AND  PIN  6 OF  SOCKET  Vl. 

L3,L4-5T.#16E.  1/4*  DIA.,  SPACED  WIRE  DIA.  ON  CTC 
IRON  CORE  FORM  PLS6.  CENTER-TO-CENTER  DIS- 
TANCE BETWEEN  LS  AND  L4  IS  3/4". 

LS-  PLATE  WINDING  ■ 6T,  #16  E,  5/16*  DIA.  1/2"  LONG 

GP/D  WINDING  I 1 T.  PLASTIC  HOOKUP  WIRE  BETWEEN 
TWO  "COLD*  END  TURNS  OF  PLATE  WINDING. 

Le-  10  T-  # 20  E,  3/6*  DIA.,  1/2"  LONG.  TAP  4T.  FROM  GNO.  END 

L7,  L8-  16T.«  20  E.  CLOSE  WOUND  ON  CTC  IRON  CORE  FORM 
PLS6.  CENTER-TO-CENTER  DISTANCE  BETWEEN  L7 
AND  L6  IS  3/4". 

L9-  Plate  winding!  9t. # i6  e,  s/ie*  dia.  s/e*  long. 

GP/O  WINDING  ••  1 T.  PLASTIC  HOOKUP  WIRE  BETWEEN 
TWO  "COLD"  END  TURNS  OF  PLATE  WINDING, 

LiOA.B,  C - MALLOPY  INDUCrUNEP,\^f  jy  TYPE,  MODEL  6303 
Ll1-  0.33JJH.  J.YY.  MtLLEP  4S66  fi-F  CHOKE. 


TRANSMITTER  SECTION.  FIGURE  17 

PLATE  WINDING!  12T.  #16,  3/4-  DIA.  3/4*  L.  BAW  3011 . 
GP/D  WINDING  • 7 T-  #16.  1 " DIA.  9/16*  L.  BtW30lS 
PLACE  PLATE  CO  I L W 1 TH I N G R I D CO  1 L . CIRCUIT 
RESONATES  TO  24-26  MC.  REGION. 

PLATE  WINDING:  18T.#16E.  1/2-OlA..  1 1/2*  L.WITH 
1/4"  SPACE  AT  CENTER  OF  WINDING.  CIRCUIT  RE- 
SONATES TO  60-64  MC.  REGION. 

ANTENNA  WINDING  - 3 T.  4F20  E.  t/2"0IA.  1/4' LONG 
AT  CENTER  OF  PLATE  WINDING. 

Ll4-PZ.4rg  WINDING'  6T,  # 16  E.  1/2"  OiA.,  1 1/2*  L.  WITH 
1/4"  SPACE  AT  CENTER  OF  WINDING.  CIRCUIT  RE- 
SONATES TO  144-146  MC.  REGION. 

GRID  WINDING'  4 T.  • 20  E.  1/2"  DIA.  1/4"  LONG 
AT  CENTER  OF  PLATE  WINDING. 

V.\h’PLATE  WINDING!  SAME  AS  Ll4,  *14  E-  WIRE. 

ANTENNA  WINDING:  3T.  #20E..  1/2"  DIA.  1/4"  LONG 
AT  CENTER  OF  PLATE  WINDING. 


poses  in  the  r-f  section  of  the  receiver  and 
every  attempt  should  be  made  to  keep  the  leads 
as  short  as  possible.  Components  associated  with 
tubes  Vs,  Vo,  and  V,  ate  mounted  between  the 
socket  pins  and  an  eight  terminal  phenolic  tie- 
point  strip  attached  to  the  side  wall  of  the 
chassis  immediately  above  the  sockets.  Trim- 
ming capacitors  Co,  C.,  and  Cs  are  mounted 
directly  to  the  terminals  of  the  TV  tuner.  A 
twelve  terminal  phenolic  strip  mounts  to  the 
tight  of  the  i-f  section  of  the  receiver  and  sup- 
ports various  components  of  these  stages. 

After  the  majority  of  components  have  been 
wired,  the  r-f  coils  of  the  receiver  ate  wound 
and  soldered  in  position.  They  should  be  ad- 
justed to  resonance  with  the  aid  of  a grid-dip 
oscillator.  The  first  tunable  i-f  circuit  (LwA-Ca) 
and  the  6BE6  grid  circuit  (LmB-Ci)  may  be 
adjusted  with  the  aid  of  the  grid-dip  oscillator 
to  covet  30  - 34  Me.  as  the  main  tuning  dial 
of  the  receiver  is  tuned  throughout  its  range. 
The  mixer  circuit  (LioC-Cs)  covers  the  range 
of  28  - 32  Me. 


Figure  22 

TEST  POWER  SUPPLY  FOR 
TRANSCEIVER 


Testing  the  Receiver  Section.  After  the  re- 
ceiver wiring  is  completed,  it  should  be  thor- 
oughly checked  for  errors  and  omissions.  Have 
another  person  check  your  wiring,  as  it  is  often 
difficult  to  find  your  own  errors.  When  you 
are  sure  of  the  circuitry  the  receiver  section  may 
be  connected  to  the  test  power  supply  shown  in 
figure  22.  This  supply  makes  use  of  the  trans- 
ceiver components  and  may  be  wired  "bread- 
board style”  for  the  duration  of  the  tests.  A 
speaker  is  connected  to  the  output  jack  of  the 
receiver,  and  a modulated  test  signal  of  the  in- 
termediate frequency  is  injected  on  pin  5 of  the 
6BE6  mixer  tube  (V«).  A very  small  capacitor 
is  used  to  couple  the  i-f  signal  into  the  receiver. 
Remove  the  6J6  oscillator  tube  (V?)  for  this 
test.  The  signal  generator  can  be  of  the  BC-221 
type  capable  of  being  accurately  set  to  a fre- 
quency half-way  between  the  crystal  frequencies 
of  the  lattice  filter.  (In  this  case,  the  mid-fre- 
quency is  2009  kc.)  Switch  Si  is  set  to  the 
"sharp”  position  and  the  tuning  slugs  of  the 
i-f  transformers  are  adjusted  for  maximum 
audio  output.  As  the  i-f  strip  is  brought  into 
alignment,  the  coupling  to  the  signal  generator 
should  be  reduced  to  prevent  overloading. 

The  6J6  mixer  tube  is  now  replaced  and  the 
signal  generator  is  tuned  to  the  first  inter- 
mediate frequency  of  34  Me.  The  tuner  is  ad- 
justed to  the  high  frequency  dial  setting  and 
oscillator  trimming  capacitor  G adjusted  until 
generator  is  heard.  R-f  padding  capacitors  G 
and  G are  adjusted  for  maximum  signal 
strength.  If  a short  antenna  is  attached  to  the 
rotor  arm  of  transfer  switch  SiA,  signals  in  the 
30  - 34  Me.  range  should  be  easily  received. 

Switch  Si  is  next  set  to  50  Me.  and  the  r-f 
amplifier  (Va)  and  mixer  tube  (Vi)  for  tjais 
range  are  placed  in  their  sockets.  Oscillator 
tuning  capacitor  G and  the  coupling  between 
coil  Lb  and  the  crystal  feedback  winding  are  ad- 
justed for  stable  operation  of  the  oscillator 
stage.  The  oscillator  may  be  monitored  in  a 


588  Low  Power  Transmitters 


THE  RADIO 


nearby  receiver  tuned  to  84  Me.  A signal  in 
the  50  - 54  Me.  range  is  now  injected  in  the 
antenna  receptacle  (J2)  of  the  receiver  section, 
and  the  adjustable  slugs  of  coils  Lt  and  Ls 
tuned  for  greatest  signal  response.  The  last 
step  is  to  adjust  the  slug  of  antenna  coil  Le. 

Switch  S7  should  be  set  to  the  144  - 148 
Me.  range  and  tubes  Vi  and  V2  are  placed  in 
their  sockets.  Oscillator  operation  is  checked 
at  114  Me.,  and  coils  L<,  Li,  and  Li  are  adjusted 
for  maximum  signal  response  in  the  2 meter 
band.  When  the  receiver  is  operating  properly 
on  both  bands,  the  transmitter  section  of  the 
transceiver  may  be  wired  and  tested. 

Wiring  the  Transmitter  Section.  The  r-f  sec- 
tion of  the  transmitter  is  wired  first.  Coupling 
coil  Lij  between  the  oscillator  and  buffer  is 
made  of  two  sections  of  coil  stock  that  are 
telescoping,  and  the  plate  coil  is  placed  within 
the  grid  coil.  Coils  Lu  and  L u are  mounted 
between  the  wafer  sections  of  Switch  Sa.  The 
50  Me.  antenna  coil  is  placed  between  the  two 
windings  of  Lia,  and  the  144  Me.  grid  coil  of 
the  2 meter  amplifier  is  placed  between  the 
dual  windings  of  coil  L14.  The  various  resistors 
associated  with  the  bandswitch  are  mounted 
between  the  terminals  of  the  switch.  The  2 
meter  plate  coil  Lu  is  fastened  to  the  terminals 
of  tuning  capacitor  C«,  which  project  through 

-inch  clearance  holes  cut  in  the  chassis.  The 
2 meter  antenna  coil  is  supported  on  two  min- 
iature ceramic  insulator  posts  mounted  adja- 
cent to  Lis.  When  the  r-f  wiring  is  completed, 
the  various  coils  may  be  adjusted  to  operating 
frequency  range  by  varying  the  spacing  between 
the  turns. 

The  audio  section  of  the  transmitter  should 
be  wired  next.  This  is  a straightforward  oper- 
ation and  comparatively  simple.  Small  shielded 
leads  are  run  from  the  second  section  of  the 
12AX7  speech  amplifier  to  the  audio  gain 
control  Ri  mounted  on  the  tear  apron  of  the 
chassis  near  the  microphone  jack.  A shielded 
lead  runs  from  the  microphone  jack  to  the 
first  section  of  the  12AX7.  The  coils  of  relays 
RYi  and  RY2  are  connected  in  series  with  the 
control  pin  of  Js,  and  are  returned  to  the  fila- 
ment circuit  of  the  transmitter. 

Testing  the  Transmitter  Section.  The  trans- 
mitter is  now  temporarily  connected  to  the 
auxiliary  power  supply  and  the  6CL6  oscillator 
tube  and  an  appropriate  crystal  are  inserted  in 
their  respective  sockets.  A test  lamp  should 
be  made  from  a 6.3  volt,  150  ma.  (brown 
bead)  pilot  lamp  attached  to  a small  loop  of 
wire.  This  test  lamp  is  coupled  loosely  to  os- 


cillator coil  Li2  and  will  light  brightly  when  the 
oscillator  stage  is  functioning.  Capacitor  Ce  is 
adjusted  for  steady  operation  of  the  oscillator. 
Bandswitch  Ss  is  now  set  to  50  Me.  and  switch 
Ss  is  set  to  the  "tune”  position  removing  plate 
and  screen  voltage  from  the  amplifier  tube. 
The  6360  is  inserted  in  the  buffer  socket  and 
meter  switch  S»  is  set  to  read  the  grid  current 
of  this  stage.  A reading  of  about  3.5  milliam- 
peres  can  be  obtained  by  proper  adjustment  of 
capacitor  Co.  The  plate  circuit  of  the  50  Me. 
stage  should  be  set  to  approximate  resonant 
frequency  with  a grid-dip  osilclator  and  a 
dummy  load  or  antenna  is  attached  to  the  50 
Me.  antenna  receptacle  J2.  Switch  Ss  is  now  set 
to  the  operate  position  and  the  push-to-talk 
relay  circuit  actuated.  Amplifier  tuning  capaci- 
tor Ci  is  adjusted  for  maximum  meter  reading 
when  the  meter  is  switched  to  the  r-f  volt- 
meter position.  Plate  current  of  the  6306  tube 
under  this  condition  is  approximately  70  milli- 
amperes.  The  actual  value  of  plate  current  may 
be  checked  by  setting  meter  switch  Ss  to  the 
"amplifier”  position  (0-  100  milliamperes ) . 

The  next  step  is  to  check  the  modulator.  In- 
sert tubes  Vis,  Vis,  and  V20  in  their  respective 
sockets.  Advance  gain  control  R2  while  speak- 
ing into  the  microphone  in  a normal  tone. 
When  checked  in  a nearby  receiver  the  modu- 
lation should  be  clear  and  crisp. 

After  the  transmitter  section  is  operating 
properly  on  50  Me,,  attention  should  be  given 
to  the  2 meter  section.  A 144  Me.  antenna  is 
attached  to  receptacle  Ji,  and  the  6360  ampli- 
fier tube  (Vii)  is  inserted  in  its  socket.  Switch 
Ss  is  placed  in  the  "tune”  position,  and  the 
coupling  between  the  grid  coil  of  this  stage  and 
buffer  plate  coil  Lh  adjusted  for  3.5  milliam- 
peres of  grid  current  to  Vn.  Switch  S.i  auto- 
matically activates  tube  Vn  when  the  switch 
is  placed  in  the  144  Me.  position.  Finally, 
switch  Ss  is  placed  in  the  "operate”  position 
and  the  tuning  (Cs)  and  loading  (Go)  controls 
are  adjusted  for  maximum  indication  on  the 
r-f  voltmeter.  Plate  current  to  the  2 meter 
stage  is  held  at  70  milliamperes  as  measured 
on  the  amplifier  position  of  meter  switch  Ss. 

The  bandchange  operation  will  now  merely 
consist  of  selecting  the  proper  transmitting 
crystal,  setting  Ss  to  the  proper  position,  and 
adjusting  the  transmitter  controls  for  maximum 
reading  of  the  r-f  voltmeter.  The  complete 
bandchanging  operation  takes  about  fifteen 
seconds. 

The  Power  When  the  transceiver  is  in  work- 
Supply  ing  order,  attention  should  be 

turned  to  the  power  supply  sec- 


HANDBOOK 


Miniature  S.S.B.  Transmitter  589 


tion  ( figure  18).  At  the  time  of  construction, 
no  power  transformer  was  available  that  would 
provide  the  desired  voltages  and  at  the  same 
time  be  capable  of  operation  from  both  a 115 
volt  a.c.  and  a d.c.  vibrator  primary  source.  It 
is  necessary,  therefore,  to  find  a power  trans- 
former that  has  suitable  high  voltage  windings 
and  to  remove  the  filament  windings,  replacing 
them  with  a special  multi-purpose  winding 
that  will  fill  the  bill.  The  Stancor  P-8158  was 
used  in  this  transceiver,  and  does  the  job  nicely. 
This  transformer  has  a 1 1 5 volt  primary  wind- 
ing, a 117  volt  secondary  winding  designed 
for  voltage  doublet  service,  and  three  6.3  volt 
filament  windings.  The  three  filament  windings 
are  on  the  outside  of  the  transformer  and  may 
be  easily  removed. 

Power  Transformer  Modification.  The  four 
transformer  bolts  should  be  removed,  exposing 
the  copper  shorting  band  of  the  transformer, 
which  is  taken  off.  Take  off  the  layers  of  cam- 
bric tape  to  expose  the  uppermost  filament 
winding.  Now,  remove  the  unused  filament 
windings,  being  careful  not  to  damage  the 
other  windings  of  the  transformer.  A new 
winding  made  of  enameled  wire  should  now 
be  placed  over  the  remaining  windings.  This 
winding  will  be  for  either  six  or  twelve  volt 
operation,  and  will  have  a voltage  tap  for  fila- 
ment operation  when  the  supply  is  run  from 
alternating  current.  During  battery  operation 
the  filaments  of  the  transceiver  will  be  run 
directly  from  the  d.c.  primary  source. 

Decide  whether  you  want  a six  or  twelve 
volt  primary  winding.  The  six  volt  winding 
will  be  made  of  #14  wire,  whereas  the  twelve 
volt  winding  will  be  made  of  #16  wire.  Each 
winding  will  have  a tap  at  the  proper  point  to 
provide  voltage  for  the  tube  filaments.  This 
new  winding  is  shown  on  transformer  Ti, 
figure  18. 

The  new  winding  will  be  put  on  in  two 
halves,  since  two  separate  windings  ate  re- 
quired for  proper  operation  of  the  "full-wave” 
vibrator.  The  six  volt  winding  consists  of  two 
separate  windings  of  14  turns  of  #14  wire. 
The  windings  are  wound  on  in  series,  with  the 
two  adjacent  leads  brought  out  as  a center  tap 
(see  terminal  B,  Ti,  figure  18).  The  6.3  volt 
filament  tap  (terminal  D,  Ti)  is  placed  ten 
turns  from  the  junction  of  the  two  windings 
(point  B).  The  windings  should  be  carefully 
taped  in  place  and  the  transformer  reassembled. 
If  twelve  volt  operation  is  desired,  two  separate 
windings  of  28  turns  of  #16  wire  are  wound 
in  series  on  the  transformer.  The  filament  tap 
for  a-c  operation  of  the  filaments  (D)  is 


placed  at  the  12.6  volt  spot  on  the  winding, 
or  twenty  turns  from  point  B.  Thus,  the  fila- 
ments work  as  a "six  volt  string”  for  6/115 
volt  operation  and  as  a "twelve  volt  string” 
for  12/115  volt  operation.  The  two  different 
arrangements  of  plug  PL-2  necessary  to  ac- 
complish this  are  shown  in  figure  18. 

Testing  the  Power  Supply.  After  transformer 
Ti  has  been  modified,  the  power  supply  should 
be  wired  in  accord  with  figure  18.  Before  it  is 
used  with  the  transceiver,  it  should  be  tested 
with  both  a.c.  and  d.c.  input  circuits.  The  high 
voltage  system  should  develop  approximately 
250  volts  across  a 1600  ohm  50  watt  resistor, 
and  a full  6.3  volts  should  be  measured  across 
a 1 ohm,  10  watt  filament  load  resistor.  The 
12  volt  configuration  should  develop  the  same 
high  voltage  under  a similar  load,  and  should 
deliver  12.6  volts  across  a 2 ohm,  10  watt  re- 
sistor. If  it  is  found  that  the  voltages  developed 
during  operation  from  the  d-c  source  are  lower 
than  the  corresponding  voltages  derived  from 
a-c  operation,  it  is  an  indication  that  the  modi- 
fied primary  winding  has  too  many  turns.  The 
transformer  should  be  removed  from  the  in- 
stallation and  one  turn  removed  from  each  half 
of  the  primary  winding.  This  will  boost  the 
d.c.  plate  voltage  about  10%.  After  the  voltage 
has  been  set,  the  power  supply  wiring  and 
switching  circuits  of  the  transceiver  may  be 
completed. 

28-4  A Miniaturired 
SSB  Transmitter 
for  14  Me. 

This  40  watt  p.e.p.  sideband  transmitter  was 
designed  to  provide  the  greatest  amount  of 
'talk  power”  for  a given  size  and  weight.  It 
was  built  by  WOAVA  expressly  for  the  pur- 
pose of  permitting  amateurs  in  far-off  coun- 
tries to  have  the  opportunity  to  experiment 
with  SSB  transmission.  The  completely  self- 
contained  transmitter  has  been  shipped  around 
the  world  by  air  freight,  and  has  been  operated 
in  such  locations  as  Canton  Island,  Fiji,  Brunei, 
Singapore,  and  other  exotic  spots. 

The  transmitter  and  power  supply  is  con- 
tained within  an  aluminum  case  measuring 
9"  X 6"  X 5"  and  weighs  less  than  seven 
pounds.  When  encased  in  a strong  cardboard 
box  it  is  completely  transportable  by  air  ex- 
press. Primary  power  source  can  be  either  115- 
or  230-volts,  50  - 60  cycles.  Air  tested  by  many 
hours  of  operation  under  conditions  of  extreme 
heat  and  humidity,  this  small  sideband  trans- 
mitter has  proven  to  be  an  effective  design 


590  Low  Power  Transmitters 


THE  RADIO 


that  may  provide  inspiration  for  a more  elab- 
orate assembly.  Tuning  adjustments  are  simple 
and  fool  proof,  and  the  transmitter  remains  in 
alignment  without  time  consuming  balancing 
of  r-f  and  audio  phase  networks. 

Transmitter  A block  diagram  of  the  circuit 
Circuitry  is  shown  in  figure  24.  Fourteen 

tubes  are  used  including  two 
voltage  regulators.  The  filter  method  of  side- 
band generation  is  employed  using  a 500  kc. 
Collins  mechanical  filter  having  a passband  of 
3.1  kc.  The  sideband  signal  is  heterodyned  to 
the  14  Me.  operating  frequency  by  two  mixer 
stages.  A voltage  doubler  type  silicon  rectifier 
power  supply  provides  high  voltage  for  trans- 
mitter operation. 

The  complete  transmitter  schematic  is  given 
in  figures  26  and  29.  Sub-miniature  tubes  are 
used  wherever  possible  to  conserve  space.  The 
audio  and  sideband  generation  section  of  the 
transmitter  is  illustrated  in  figure  26.  The  heart 
of  this  section  is  the  Crosby  triple  triode  mod- 
ulator stage  (VsA,  VaA,  VaB).  The  first 
triode  serves  as  a cathode  follower  audio  stage 
and  is  driven  by  a sub-miniature  5702  resis- 
tance coupled  audio  amplifier  Vi.  The  second 
triode  is  also  a cathode  follower  (VaA)  and 
is  excited  by  the  mixing  oscillator  stage  V<A, 
B.  The  cathode  follower  output  voltage  is  de- 
veloped across  a common  cathode  resistor  and 
is  injected  into  the  cathode  of  a grounded  grid 
mixer  tube  VaB.  A double  sideband  A-M 
signal  is  developed  in  the  plate  circuit  of  the 
mixer  tube.  The  phase  of  the  signal  at  the 
plate  of  mixer  VaB  is  opposite  to  that  of  a 
signal  injected  on  the  control  grid  of  the  tube. 
Cancellation  of  the  carrier  is  therefore  possible 
by  injecting  the  r-f  carrier  on  this  grid  via 


the  "carrier  null”  potentiometer  Ra.  For  maxi- 
mum carrier  elimination  capacitor  Ci  is  ad- 
justed to  neutrali2e  the  capacity  feed-through 
which  may  occur  between  VaA  and  VaB. 
With  proper  adjustment  of  Q and  Ra  a car- 
rier reduction  of  greater  than  40  decibels  be- 
low maximum  SSB  signal  may  be  obtained 
with  stability. 

The  mixing  carrier  is  generated  at  approx- 
imately 498  kc.,  the  frequency  of  the  "20 
db  attenuation  point”  of  the  mechanical  filter. 
The  carrier  is  coupled  into  the  Crosby  modu- 
lator through  cathode  follower  VaB.  A single 
sideband  suppressed  carrier  signal  appears  at 
the  output  terminals  of  the  mechanical  filter 
and  is  amplified  by  a single  5702  pentode 
stage  (Vj),  which  drives  a cathode  coupled 
6BG7  mixer  (Vo).  A crystal  oscillator  (Vw) 
operating  on  4002  kc.  provides  mixing  voltage 
to  deliver  a 4.5  Me.  sideband  signal  in  the 
output  circuit  of  the  mixer  tube.  Two  double 
tuned  transformers  (Tz  and  Ts)  provide  ade- 
quate supression  of  the  mixer  voltage.  Further 
signal  amplification  is  obtained  at  4.5  Me.  by 
virtue  of  a 6BG7  cascode  r-f  amplifier  stage, 
V,. 

At  this  point  the  signal  is  mixed  in  a second 
cathode  coupled  mixer  stage  (Vs)  and  hetero- 
dyned to  the  14  Me.  amateur  band.  The  mix- 
ing oscillator  operates  in  the  9-7  - 9-8  Me.  re- 
gion and  uses  a 5744  triode  (Vn)  in  a Pierce 
circuit.  The  14  Me.  SSB  signal  is  developed 
across  a tuned  circuit  in  the  plate  of  the  mixer 
tube.  The  signal  level  at  this  point  is  quite 
low,  and  a second  cascode  r-f  amplifier  (Vs) 
is  employed  to  boost  the  signal  to  a level  high 
enough  to  drive  a class  ABi  tetrode  linear  am- 
plifier. The  RCA  6524  dual  beam  amplifier 
tube  requires  only  33  volts  peak  drive  signal 
to  develop  50  watts  p.e.p.,  so  it  was  chosen  for 
the  final  r-f  amphfier  stage.  With  the  two  sec- 
tions of  the  tube  connected  in  parallel  and  475 
volts  applied  to  the  plates  approximately  40 
watts  p.e.p.  can  be  obtained  from  the  tube. 
For  maximum  stability  and  freedom  from  para- 
sitic oscillations,  suppressors  are  placed  in  the 
grid  and  plate  leads  of  the  tube  and  the  stage 
is  neutralized. 

A simple  pi-section  output  circuit  is  em- 


Figure  23 

SINGLE  SIDEBAND  IN  A SMALL  PACKAGE! 

A complete  filter-type  14  Me.  single  sideband  trans- 
mitter capable  of  40  watts  p.e.p.  and  less  than  one-half 
cubic  foot  in  size  is  described  in  this  section.  Com- 
pletely self  contained,  this  seven  pound  sideband  pack- 
age is  well  suited  for  portable  or  mobile  operation.  Pri- 
mary power  source  is  115/230  volts,  50-60  cycles.  Sili- 
con rectifiers  are  used  in  high  voltage  supply  and  a 
6524  is  employed  as  high  level  linear  amplifier. 


HANDBOOK 


Miniature  S.S.B.  Transmitter  591 


AUDIO  CATHODE  PRODUCT  t-F  SECOND  4.5MC.  THIRD  FIRST14MC  SEC0ND14MC. 

AMP.  FOLLOWER  MODULATOR  AMPLIFIER  MIXER  AMPLIFIER  MIXER  AMPLIFIER  AMPLIFIER 


498  KC  4002  KC.  9.7-9. BMC. 


116/2S0V, 

50-60 


LAYOUT,  CHASSIS ♦ 1,  SIDE*  1 


LAYOUT,  CHASSIS  4t1,  FRONT 


LAYOUT,  CHASSIS  « 1,  SIDE  » 2 


Figure  24 

BLOCK  DIAGRAM  AND  PARTS  LAYOUT  OF  14  MC.  SIDEBAND  TRANSMITTER 

fourteen  tuber  (Intluding  two  voltage  regulators)  are  used  in  this  compact  transmitter.  A "Crosby" 
modulator  and  500  kc.  Collins  Mechanical  filter  are  employed  for  sideband  generation,  first  frequency 
conversion  is  to  4 Me.,  and  second  frequency  conversion  is  to  14  Me. 

Layout  of  major  components  on  both  sides  and  front  of  special  chassis  is  shown  in  the  illustration. 


Figure  25 

OBLIQUE  VIEW  OF 
AMPLIFIER  AND 
EXCITER  CHASSIS 

The  power  amplifier 

chassis  is  at  the  left.  The 
plate  coil,  plate  tuning 
capacitor  and  loading 
switch  are  in  the  fore~ 
ground.  Side  # 7 of  the 
exciter  is  shown.  The  in~ 
ter-ehassis  area  is  cov- 
ered with  perforated 

metal.  The  tip  jacks  for 

the  receiver  disabling 

circuit  are  on  the  right 
edge  of  the  chassis.  Cor- 
ners of  the  mounting 
base  are  rounded  to  per- 
mit close  fit  with  round- 
ed corners  of  drawn 
aluminum  case. 


592  Low  Power  Transmitters 


ployed  with  the  linear  amplifier  stage.  In  order 
to  conserve  space,  the  usual  variable  capacitor 
is  omitted  from  the  output  side  of  the  network 
and  a rotary  switch  and  selection  of  fixed  ca- 
pacitors is  employed  in  its  place.  Proper  choice 
of  capacitors  permits  the  transmitter  to  be 
matched  to  52-  or  72-ohm  transmission  lines 
having  s.w.r.  values  as  great  as  2.5/1. 

The  problem  of  obtaining  the  necessary  d-c 
power  from  a supply  that  would  fit  within 
the  small  cabinet  space  was  solved  by  employ- 
ing a bridge  rectifier  using  the  new  Sarkes- 


Tarzian  silicon  rectifiers.  The  forward  voltage 
drop  of  these  rectifiers  is  of  the  order  of  only 
1.5  volts  and  the  back  resistance  is  extremely 
high.  Because  of  the  very  small  size,  a great 
number  of  rectifiers  may  be  placed  in  a small 
area  (figure  33).  Care  must  be  taken  with 
these  rectifiers  not  to  overload  them  with  an 
accidental  short  circuit.  Peak  currents  are  cri- 
tical because  the  small  mass  of  the  rectifier 
element  will  heat  instantaneously  and  could 
conceivably  reach  failure  temperature  within 
a time  lapse  of  a few  microseconds.  The  use  of 


SIDE  * 1 OF  CHASSIS*  1 


SCHEMATIC,  LOW  FREQUENCY  SECTION  AND  POWER  SUPPLY  OF  SSB  TRANSMITTER 

DL: — Time  delay  relay,  45  seconds.  Amperite  6N045-T.  Normally  open, 

Ti — 455  kc.  miniature  i-f  transformer.  J.  W.  Miller  I2-C9.  Remove  turns  from  windings  to  resonate  at 
500  kc. 

Ti — ISO  volt,  125  ma.,  500  vo/1,  725  ma.,  6.3  volt,  4 amp.  Two  115  volt,  50-60  cycle  primary  windings. 

Walgren  3266.  Walgren  Electric  Mfg.  Co.,  Pasadena,  Calif. 

CHi,  CHi — 2 henry  at  130  ma.  Stancor  C-2303 

SRi — Four  silicon  rectifiers.  Max.  inverse  volts  ~ 280.  Sorkes-Tarxian  M-500  or  1N1084 

SRi — Eight  silicon  rectifiers,  two  in  series  tor  each  leg.  Max.  Inverse  volts  — 280.  Sarkes-Tarzian  M-500 
or  7NI084 

MF — Collings  Mechanical  Filter,  3.1  kc.  bandwidth,  500  kc.  center  frequency.  Type  SOOB-31 
Xi — Approximately  498  kc.  Frequency  chosen  to  place  carrier  oscillator  at  "20  db.  down"  point  on  filter 
curve  (see  filter  data  sheet). 

RFC — 2.5  mh.  miniature  r-f  choke.  Millen  J300-2500. 

Si — See  figure  29 


Figure  27 

INTERIOR  OF  LOW  FREQUENCY 
PORTION  OF  FILTER  TYPE  SSB 
TRANSMITTER 

500  kc.  lilier  and  low  frequency  components 
are  on  the  bottom  deck  of  the  chassis.  Smalt 
components  are  mounted  between  socket  pins 
and  miniature  tie-point  terminals.  Interior 
shield  isolates  low  frequency  section  from 
conversion  oscillators  and  4 Me.  amplifier 
stages.  The  "twenty  meter"  conversion  crys- 
tal Xs  projects  through  the  front  panel  of 
the  transmitter  (left).  Note  that  transformers 
are  mounted  below  chassis  level  to  bring 
overall  height  even  with  that  of  miniature 
tubes.  6BK7A  amplifier  tube  is  also  sub- 
mounted. 

Power  leads  that  pass  through  Inter-stage 
shielding  are  routed  through  feed-thru  type 
ceramic  insulators  (Centralab  type  FT-IOOO). 


an  auxiliary  supply  for  tuning  operations, 
therefore,  is  highly  recommended. 

Transmitter  Layout  The  transmitter  is  built 
and  Assembly  in  three  sections  which 

ate  held  together  by  a 
common  front  panel.  The  exciter  ( figure  27 ) , 
the  linear  amplifier  ( figure  31),  and  the  power 
supply  (figure  33)  make  up  these  three  sec- 
tions. When  the  three  sections  are  placed  in 
position  they  appear  as  in  figure  32.  A rear 
view  of  the  complete  transmitter  (minus  the 
case)  is  shown  in  figure  28.  The  sections  are 
bolted  together  and  the  complete  assembly  is 
fastened  to  the  front  panel.  Because  space  is  at 
a premium  in  such  a configuration,  each  chas- 
sis is  individually  designed  and  shaped  to  fit 
the  unusual  layout.  A sketch  of  the  exciter 
chassis  is  shown  in  figure  24.  It  is  formed 
from  two  sides  and  a base.  The  base  mounts 
in  a vertical  position  in  the  completed  as- 
sembly (see  figure  28).  Side  #1  of  the  ex- 
citer chassis  contains  the  low  frequency  portion 
of  the  exciter  and  can  be  seen  in  figure  26,  and 


the  actual  placement  of  major  components  is 
shown  in  figures  25  and  27.  The  mechanical 
filter  occupies  the  upper  right  portion  of  the 
deck.  Input  and  output  circuits  of  the  filter 
are  isolated  by  a shield  partition  that  passes 
across  the  midsection  of  the  filter.  This  shield 
also  braces  side  #1  to  the  interior  full- 
length  shield  seen  in  figure  27.  Power  leads 
from  this  portion  of  the  exciter  pass  through 
.001  ggfd.  Centralab  type  FT  ceramic  feed- 
through capacitors.  These  capacitors  are  em- 
ployed wherever  power  leads  pass  through  an 
interstage  shield.  Note  that  coupling  trans- 
former Ti  is  submounted  to  bring  its  height 
in  line  with  that  of  the  sub-miniature  tubes- 
All  components  are  mounted  on  this  side  of 
the  chassis  and  it  is  wired  before  it  is  attached 
to  the  base. 

Side  #2  of  the  exciter  chassis  is  attached 
to  the  base,  and  is  further  braced  to  side  #1 
by  an  end  plate  and  an  interstage  shield.  The 
6BK7A  socket,  and  transformers  Ts  and  Ts 
are  submounted  to  bring  their  height  down  to 
that  of  the  sub-miniature  tubes.  The  compon- 


Figure  28 

REAR  VIEW  OF  ASSEMBLED  SSB 
TRANSMITTER 

The  three  sub-assemblies  of  the  transmitter 
ore  fastened  together  and  bolted  to  the  front 
panel.  At  the  left  is  the  power  supply  section. 
The  o.c.  power  receptacle  is  at  the  bottom 
of  the  assembly  with  the  two  filter  chokes 
at  the  left.  Above  the  chokes  is  the  bank  of 
silicon  rectifiers  with  the  two  voltage  regula- 
tor tubes  and  dropping  resistor  at  the  top  of 
the  supply. 

The  linear  amplifier  comprises  the  center  sec- 
tion of  the  assembly.  Antenna  and  receiver 
coaxial  receptacles  are  mounted  on  a small 
bracket  bolted  between  the  outer  sections. 
Plate  coil  is  visible  at  the  top  of  the  chassis 
with  adjustable  loading  switch  to  the  right. 

The  right  section  of  the  assembly  is  the  ex- 
citer/ shown  in  detail  in  figure  27. 


594  Low  Power  Transmitters 


THE  RADIO 


SIDE  » 2 OF  CHASSIS  # 1 

V6  V7  Vs  V9 


Via  CHASSIS#a 

6524  PC 


Tf/  Tj — 4.5  Me.  intersfoge  iransformer.  J.  W.  Miller  6204 

Li — 20  turns  #22  e.,  ^4"  diam..  3/q"  long  on  ceramic  form  with  iron  slug.  J.  Yf.  Miller  4504 
Lt,  Ls — 75  turns  #78,  diam,  /ong  on  polystYfene  form.  Link  ~ 2 turns  hookup  wire. 

Li — 72  turns#  76,  1^/4"  diam.,  V/2"  long 

Cl — 50  MP-fd.  ceramic  trimmer.  Centralab  822-AN 

Ct — 50  pjufd.  Johnson  50K10 

C3 — 9 fififd.  Johnson  9M7  7 

Ci — 50  /x/j-fd.  Johnson  50L1S 

Cs — 75  ppfd.  fixed  capacitor.  Switch  5/  adds  seven  33  p/ifd.  capacitors  in  succession.  (El-Menco  type 
CM-79  or  Centralab  TCZ‘33). 

5iA-E — 5 pole,  3 position.  Centralab  PA‘2015 

5s — Centralab  P-121  Index  Assembly  with  PIS  progressive  shorting  deck. 

Xs — 4002  kc.  Precision  Crystal  Lab.,  Santa  Monica,  Calif. 

Xs — Frequency  — 20  meter  frequency  minus  4500  kc. 

PC — Parasitic  choke.  52  ohm,  1 watt  resistor  wound  with  6 tarns  # 76  wire. 

RFC — 21/2  'n/i.  miniature  r-f  choke.  Millen  J300-2500 
M — 0-10  d.c.  milliammeter. 

Bi — 37.5  volts.  Burgess  XX22  plus  two  type  Z flashlight  cells 
SHi — 750  mo.  shunt.  100  ohms 
RFCi — Ohmite  Z-14  r-f  choke 


ents  are  mounted  and  wired  before  the  side  is 
attached  to  the  base.  The  base  contains  no  wir- 
ing, except  for  two  leads  to  the  receiver  dis- 
abling jacks  J4  and  Js. 

Amplifier  chassis  #2  occupies  the  center 


portion  of  the  transmitter  assembly,  and  can 
be  seen  in  figures  23,  25,  and  30.  The  tube 
socket  and  major  parts  are  mounted  on  a flat 
plate,  with  the  grid  circuit  components  en- 
closed in  a "step”  shaped  box.  This  shield  is 


HANDBOOK 


Miniature  S.S.B.  Transmitter  595 


Figure  30 

LINEAR  AMPLIFIER 
AND  EXCITER 
SECTIONS  OF  SSB 
TRANSMITTER 
Left:  Linear  amplifier  Is 
constructed  upon  on 
aluminum  sheet  with 
grid  circuit  enclosed  by 
small  L-shaped  shield 
box.  Low  impedance  link 
leads  from  exciter  pass 
through  grommet  in  box. 
The  coaxial  output  ca- 
ble passes  through  chas- 
sis hole  to  changeover 
switch.  Space  to  the  right 
of  grid  enclosure  is  oc- 
cupied by  meter  and  time 
delay  relay. 

Right:  Front  view  of  as- 
sembled exciter  chassis 
shows  main  panel  con- 
trols. Chassis  mounts  in 
vertical  position  in  final 
assembly.  ^'Twenty  me- 
ter*' conversion  crystal 
socket  is  at  top,  left,  of 
front  of  assembly. 


clearly  visible  in  figure  32.  The  clearance  space 
provided  by  the  "step"  is  occupied  by  the 
"SSB-  standby-CW”  switch  Si  which  is  mount- 
ed to  the  panel  in  close  proximity  to  the  grid 
circuit  components  of  the  amplifier  stage.  The 
important  leads  to  the  control  switch  pass  out 
of  the  grid  compartment  through  ceramic  feed 
through  capacitors  mounted  on  the  wall  of  the 
"step”  shield.  The  coaxial  r-f  output  lead  and 
the  grid  bias  lead  also  pass  through  this  shield 
and  may  be  seen  in  figure  32. 

Figure  31  provides  a close-up  of  the  in- 
terior section  of  the  amplifier  chassis.  Grid 
coil  La  is  affixed  to  a polystyrene  insulator  and 
is  mounted  to  the  side  of  the  chassis-box.  Neu- 
tralizing capacitor  Ca  is  placed  on  a small  poly- 
styrene plate  adjacent  to  the  6524  amplifier 
tube.  Plate  r-f  choke  is  next  to  Ca,  and  the 
5KV  plate  blocking  capacitor  is  supported  be- 
tween the  top  end  of  the  choke  and  the  stator 
of  tuning  capacitor  G.  The  plate  inductor  L» 


occupies  the  far  end  of  the  chassis.  Pi-network 
switch  S:  is  placed  next  to  the  tuning  capaci- 
tor, and  the  various  padding  capacitors  are 
mounted  directly  on  the  back  of  the  switch 
section. 

Power  supply  chassis  #3  is  attached  to  the 
front  panel,  and  the  weight  of  the  supply  is 
supported  by  a sheet  metal  screw  run  through 
the  rear  of  the  case  into  the  supply  chassis 
after  final  assembly.  The  power  supply  com- 
ponents are  mounted  upon  a phenolic  board 
which  in  turn  is  fastened  to  the  aluminum 
chassis  frame.  The  silicon  rectifiers  are  held  in 
position  by  fuse  clips  mounted  on  the  boards 
and  are  placed  so  that  they  obtain  the  maxi- 
mum possible  ventilation.  The  primary  power 
plug  is  placed  on  the  rear  of  the  supply  chas- 
sis and  projects  through  a hole  cut  in  the  case 
of  the  transmitter.  The  time  delay  relay  DL-1 
is  mounted  on  the  side  of  the  amplifier  chassis 
and  may  be  seen  in  figure  31.  The  small  bias 
battery  for  the  amplifier  stage  fits  below  this 
space. 

Transmitter  It  is  important  that  no  low 
Wiring  frequency  energy  pass  around 

the  mechanical  filter  as  spurious 


Figure  31 

INTERIOR  VIEW  OF  LINEAR 
AMPLIFIER  GRID  COMPARTMENT 

7/me  delay  relay  is  mounted  on  side  of  grid 
box,  occupying  space  between  the  panel  meter 
and  plate  tank  assembly.  The  grid  coil  and 
tuning  capacitor  can  be  seen  below  the  tube 
socket.  Neutralizing  capacitor  is  made  from 
Johnson  30M8  with  every  third  plate  removed, 
but  capacitor  listed  in  parts  list  of  figure  29 
is  satisfactory. 


leakage  will  deteriorate  the  sideband  and  car- 
rier suppression  to  a great  degree.  Power  leads 
are  therefore  bypassed  with  feed  through  type 
ceramic  capacitors  at  each  partition.  Several 
circuits  (ViB,  V2,  V.i,  V?,  and  V»)  are  of 
the  "hot  cathode”  type  in  which  r-f  voltage 
appears  on  the  cathode  of  the  tube.  It  is  very 
necessary,  therefore,  to  bypass  the  filament 
lead  of  these  tubes  to  prevent  a signal  leakage 
path  along  the  common  filament  wiring. 

Bypass  capacitors  are  placed  directly  at  the 
tube  socket  pins  to  conserve  space  and  all 
wiring  is  short  and  direct.  Coil  Li  (14  Me. 
mixer  plate  coil)  is  mounted  within  the  chassis 
assembly  on  a small  bracket  and  a short  ex- 
tension shaft  is  soldered  to  the  slug  to  permit 
circuit  adjustment  from  the  front  panel.  All 
wiring  must  be  checked  for  opens,  shorts,  trans- 
positions, and  accidental  grounds  before  power 
is  applied  to  the  transmitter. 

Testing  the  The  transmitter  should  be  test- 
Transmitter  ed  in  sections  before  it  is  as- 
sembled in  the  case.  The  use 
of  an  auxiliary  power  supply  is  recommended. 
The  exciter  section  should  be  tested  first.  A 
one  kilocycle  audio  signal  of  low  harmonic 
content  is  applied  to  the  microphone  jack  Ji. 
"Carrier  insert”  potentiometer  R2  is  set  at  zero 
(ground  end)  and  the  antenna  of  a receiver 
capable  of  tuning  to  500  kc.  is  attached  to 
point  A,  figure  26.  Before  the  audio  signal  is 
applied,  "carrier  null”  potentiometer  R;i  and 
neutralizing  capacitor  Ct  are  adjusted  for  min- 
imum signal  at  the  crystal  frequency  of  498 
kc.  When  the  audio  level  is  advanced  an  un- 
modulated carrier  should  be  heard  in  the  re- 
ceiver. The  frequency  of  the  carrier  will  be 


Figure  32 

COMPLETE  R-F  ASSEMBLY  OF  20 
METER  SSB  TRANSMITTER 

The  two  r-f  units  are  bolted  together  in  this 
view.  The  power  supply  chassis  mounts  on  the 
right-hand  edge  of  this  assembly.  Bias  bat- 
teries fit  in  the  lower  area  of  the  center  unit, 
white  plate  meter  fits  into  upper  space  in 
front  of  time  delay  relay.  Compare  this  view 
with  figure  23.  This  photo  is  taken  from  the 
right  front,  looking  upward  at  the  assembly. 

the  crystal  frequency  plus  the  frequency  of 
the  audio  signal,  (i.e.;  a 2 kc.  audio  signal 
will  produce  a carrier  frequency  of  500  kc.). 
The  carrier  may  be  observed  on  an  oscilloscope 
as  described  in  chapter  17  and  audio  signal 
level  and  null  adjustments  are  varied  to  reduce 
the  residual  "ripple”  modulation  of  the  car- 
rier. 

Conversion  crystal  X;  and  tubes  Vs,  V7,  Vs, 
and  Vio  are  inserted  in  the  proper  sockets  and 
the  receiver  is  tuned  to  4.5  Me.  and  coupled 
to  the  secondary  winding  of  Ti.  Transformer 
Ti,  T:,  and  Ts  are  adjusted  for  maximum  sig- 


Figure  33 

COMPACT  SILICON  POWER  SUPPLY 
RUNS  ENTIRE  TRANSMITTER 
Miniature  silicon  rectifiers  permit  the  com- 
plete power  supply  to  be  built  in  extremely 
small  space.  Rectifiers  are  mounted  in  fuse 
clips  bolted  to  phenolic  board.  Power  trans- 
former occupies  bottom  area  of  assembly  and 
is  fastened  to  aluminum  assembly  plate  by 
the  four  bolts  in  the  foreground.  Primary 
power  plug  can  be  seen  at  top  of  unit,  with 
the  two  filter  chokes  mounted  back-to-back 
on  an  arm  of  the  assembly  plate.  Voltage  reg- 
ulator tubes  and  dropping  resistors  are  in  the 
foreground. 


Figure  34 

"DUPLEX"  TRANSMITTER-RECEIVER 
FOR  220  MC.  RADIO  LINK 

The  width  of  the  amateur  220  Me.  band  per- 
mits simuitaneous  duplex  transmission  between 
two  remote  points.  This  compact  YHF  package 
contains  a complete  220  Me.  '^Radio  Link.**  A 
crystal  controlled  transmitter  operates  on 
224.6  Me.  The  receiver  operates  at  a fre- 
quency of  220.7  Me.  Signal  from  the  trans- 
mitter acts  as  local  oscillator  for  the  i-f  sig- 
nal of  4.5  Me. 

The  complete  station  is  housed  in  single  steel 
cabinet.  Transmitter  section  is  at  left  with 
crystal  mounted  on  panel.  Receiver  section  is 
at  right  with  power  supply  occupying  lower 
portion  of  cabinet.  Transmitter  may  be  tone 
modulated  for  i.e.w.  transmission. 


nal  level  at  4.5  Me.  If  a broadband  oscilloscope, 
such  as  the  Heathkit  0-1 1 is  at  hand,  the  signal 
may  be  observed  visually  and  the  audio  level 
may  be  set  for  minimum  carrier  modulation. 
When  monitored  in  the  receiver  the  4.5  Me. 
signal  should  be  a pure  carrier  with  little  or 
no  tone  modulation.  When  audio  gain  con- 
trol Ri  is  retarded,  the  signal  should  gradually 
weaken  and  disappear.  After  satisfactory  oper- 
ation is  obtained  at  this  frequency,  crystal  X3 


and  tubes  Va  and  Vn  are  placed  in  their  sockets. 
Capacitor  Ci  (figure  29)  is  tuned  for  maxi- 
mum SSB  signal  in  the  14  Me.  band.  The  in- 
terstage coupling  transformers  are  peaked  for 
maximum  sideband  signal  and  the  receiver 
may  be  used  for  monitoring  purposes.  Voice 
modulation  of  the  transmitter  should  be  sharp 
and  clean.  The  carrier  may  be  inserted  for 
test  purposes  by  advancing  potentiometer  R2. 

The  next  step  is  to  check  the  operation  of 


Seventeen  tubes  ore  employed  in  the  VHF  station.  The  receiver  has  two  grounded  grid  r-f  stages  for 
maximum  sensitivity.  Local  oscillator  injection  is  provided  from  the  transmitter.  Three  4.5  Me  i-f  stages 
provide  excellent  gain  and  adequate  selectivity.  Transmitter  is  crystal  controlled  from  25  Me.  crystal 
and  is  plate  modulated.  Power  suppfy  provides  300  volts  at  210  milliamperes.  Transmitter  draws  80 
milliamperes,  receiver  draws  75  milliamperes,  and  modulator  draws  approximately  45  milliamperes  under 
700%  modulation.  Speech  system  is  designated  to  be  used  with  high  gain  mobile-type  carbon  microphone. 


598  Low  Power  Transmitters 


THE  RADIO 


Figure  36 

REMOVAL  OF  FRONT  PANEL  SHOWS 
PLACEMENT  OF  R-F  CHASSIS 

The  receiver  and  transmitter  sections  are  bolt- 
ed  together  to  form  a single  unit  which  sits 
near  the  top  edges  of  the  power  transformer 
and  modulator  chassis.  Power  supply  chassis 
is  bolted  to  cabinet  *'on  edge"  along  left 
hand  side,  and  modulator  chassis  is  attached 
in  the  same  fashion  along  the  right  hand  side 
of  the  cabinet.  Ventilation  holes  are  drilled 
along  upper  edge  of  rear  portion  of  cabinet. 


the  linear  amplifier  stage.  Bias  voltage  should 
be  applied  to  the  6524  and  meter  switch  Sa 
set  to  the  grid  position.  With  full  carrier  in- 
sertion, one  or  two  milliamperes  of  grid  cur- 
rent will  flow.  The  stage  is  neutralized  by 
adjusting  capacitor  G for  minimum  grid  cur- 
rent fluctuation  as  the  plate  circuit  is  tuned 
through  resonance.  A dummy  load  is  then  at- 
tached to  the  antenna  receptacle  (Js)  and  the 
amplifier  stage  tuned  in  the  usual  manner. 
Make  sure  that  plate  voltage  is  never  removed 
when  excitation  is  applied  to  the  tube  as  the 
screen  current  will  increase  to  such  a value 
as  to  endanger  the  tube.  Always  remove  the 
screen  lead  when  you  remove  the  plate  voltage. 
When  the  carrier  is  nulled  out,  the  resting 
plate  current  of  the  linear  amplifier  will  be 
about  20  ma.,  rising  to  about  100  ma.  under 
voice  peaks.  No  grid  current  should  be  indi- 
cated under  these  conditions. 


28-5  A Duplex 

Transmifter-Receiver 
for  220  Me. 

Duplex  operation  is  permitted  in  the  220 
Me.  amateur  band  and  opens  interesting  pos- 
sibilities for  unusual  and  novel  forms  of  equip- 
ment. This  class  of  operation  consists  of  two 
one-way  communication  links  separated  in  fre- 
quency from  each  other  sufficiently  to  permit 
interference-free  operation.  The  width  of  the 
220  Me.  band  permits  placing  one  link  at  the 
low  frequency  end  of  the  band  and  the  other 
link  near  the  high  frequency  end  of  the  band 
without  the  danger  of  excessive  interference 
between  the  links. 

Shown  in  figure  35  is  the  block  diagram  of 
a transmitter-receiver  unit  designed  to  serve 
as  one  end  of  a typical  duplex  communication 
link.  Transmission  takes  place  on  224.6  Me., 
and  reception  takes  place  on  220.1  Me.  The 
frequency  separation  between  the  two  links  is 
4.5  Me.  At  the  opposite  end  of  the  link,  trans- 
mission takes  place  on  220.1  Me.,  and  recep- 
tion takes  place  on  224.6  Me. 

The  complete  duplex  station  is  built  within 
a single  cabinet  and  employs  a common  power 
supply.  Separate  antennas  are  used  for  trans- 
mission and  reception,  and  communication  is 
maintained  in  the  same  manner  as  in  the  case 
of  a "land-line,”  that  is,  simultaneous  recep- 
tion and  transmission  are  possible. 

Transmitter-Receiver  The  schematic  of  the 

Circuitry  transmitter  - receiver 

is  given  in  figure  37 
and  figure  38.  The  receiver  section  employs 
eight  tubes  in  a superheterodyne  circuit  having 
two  grounded  grid  r-f  amplifier  stages  (figure 
35).  The  intermediate  frequency  of  the  re- 
ceiver is  4.5  Me.  and  three  stages  of  i-f  ampli- 
fication are  used.  Since  the  frequency  separa- 
tion of  the  two  links  is  4.5  Me.  it  is  feasible  to 
employ  the  signal  from  the  transmitter  portion 
of  the  unit  as  the  injection  frequency  for  the 
first  mixer  of  the  receiver.  Thus  with  the  use 
of  two  properly  chosen  transmitter  crystals 
(one  at  each  end  of  the  duplex  link)  the  two 
transmitter-receivers  are  locked  on  frequency 
and  tuning  of  the  receivers  is  unnecessary.  The 
transmission  frequency  at  each  end  of  the  cir- 
cuit controls  the  frequency  of  reception  of  the 
receiver  portion  of  the  transmitter-receiver. 

Two  6AJ4  tubes  are  used  in  the  receiver  as 
grounded  grid  r-f  amplifier  stages.  The  gain 
per  stage  is  quite  low  but  is  sufficient  to  over- 
come the  noise  level  of  the  mixer  tube  (Va). 
R-f  energy  from  the  transmitter  section  pro- 


(2^,994  KC  ) 

[rfcTS  1 


-XJ  7 
',/-r',A4-  I 


TRANSMITTER  SECTION 
Vio 

(75A#C.)  12AT7  (225MC.} 


1 

47  1 

1 

— rr 

L2  T33K 

! L3  L4 

C4,4  /-'•• 

. II  LW/27X\ 

1 

2/  X 

- J 

j 

(225  MC.) 

Ls  Le  Ji,  Tx  ANT. 


La. 7 K 

r 1N34  ,0K 


"HM 

^ gRFCI 


3 .001 


2.7K 


^ :£-L  - .001  .001  500  I 

^ ’0  — ”c“ 


TO  METpR 
(0-7  DC  MA.) 


TRANSMITTER  COIL  TABLE 

L 1-  26T.,  1/2*  DIA.,  3/4"  LONG. 

Ua  - 7 T.,  1/2"  OIA,,  1/2*  LONG. 

U3-  2 T..  1/2"  DIA,,  1/4"  LONG. 

L4-  2T.,  3/8"  DIA.,  OVER  L3. 
L5-2"L0NG.  1"WlDe.  #10  W(R£ 

Le  - 1 T.  #12,  1 1/2"  OIA.  OVER  LS- 
L7-  1 6T.#16,  1/8" DIA.,  3/4"  LONG. 


Vi 

L8  6AJ4  R'F  I 

RXANT/^  J|^3,4,6,9 

^Jrlc£  7 
1X2  Jit 


NOTES 

-ALL  RESISTORS  1/2  WATT  UNLESS  OTHERWISE 
NOTED. 


V2 

6AJ4  R-F 


RECEIVER  SECTION 

V3  (4.5A/C.) 

LiO  Lit  6AJ4  MIXER  Ti 


51  1.3.4.6.9  L 

§ 

^cio.4^ 

!j\8 


m. 


^ F 

'Tool  ro'ti 


i 

PIN4,  PIN3, 

PLi  PLi 


V4 

6BA6  I-F 


V5 

6BA6  i-F 


RECEIVER  COIL  TABLE 

(ALL  COILS  1/4-  DIA.) 

L6-3T.#  16, 1/2"  L.,  ANT.  TAPI  1/2,CATH.1T. 
L9-  3 T.#  16,  3/6"  LONG,  TAP  1 1/2  T. 
LtO-3T.#16,  3/8"  LONG,  TAP  2 1/4  T. 
Ll1-2T.4kl6,  3/6"  LONG- 


6BA6  I-F  T4 


V7 

6T8  OET/ANL 


•00m  i r 1 1-2r  500670^ 

i.  r4  tjrif 


Figure  37 

schematic,  r-f  section  of  transmitter-receiver 

TRANSMITTER  SECTION  RECEIVER  SECTION 

Cl,  Cl.  Cl.  Ce — 10-10  iitifd,  Johnson  11MB11  Cs.  C,,  Cw,  C,,— 4 /iiifd.  Erie  3139-E 

RECi—Ohmite  Z-23S  r-f  choke 

Cr — 10  ^fifd.  Johnson  9M1 1 _ _ ...  . ....  .....  .... 

RFC, Ohmite  Z-28  r-f  choke  '"terstage  i-f  transformer.  J.  W.  Miller  1466 

RFCi — Ohmite  Z-235  r-f  choke  PLi — 6 coirtoct  receptacle.  Cinch-Jones  P-306-AB 

Xi — 24,944  kc.  crystal.  Precision  Crystal  Lab..  Santa  Monica.  Calif. 


220-Mc.  Duplex  599 


600  Low  Power  Transmitters 


Vi4  Via 

IZAU?  T6  12AX7  T? 


Figure  38 


SCHEMATIC  AUDIO  AND  POWER  SUPPLY  SECTIONS  OF  TRANSMITTER-RECEIVER 

Ts,  Ts — 70/C  pri.,  4 ohm  sec.  Stancor  A-3879 

Te — 70#C  pri.,  Pri.  to  y2  see.  = 2:7.  Stancor  A-47I3 

Tt — 70  pri.,  8K  sec.  Stancor  A-3845 

Ts 360-0-360  Yolts  at  200  ma„  5 v.  at  6 amp.,  6.3  v.  at  9 amp.  Stancor  P-8351 

CHi — 2 henry  at  200  ma.  Stancor  C-2325 
B — Miniature  blower  motor  and  fan,  7 75  volts 


Figure  39 

REAR  VIEW,  R-F  DECKS  OF 
TRANSMITTER-RECEIVER 

Receiver  deck  is  at  the  left  and 
transmitter  deck  is  at  the  right. 
The  two  chassis  are  bolted  to- 
gether to  form  one  unit.  Across 
the  back  lip  (left  to  right)  are; 
Receiver  antenna  receptacle, 
pwoer  plug  PLi,  meter  terminals, 
transmitter  antenna  receptacle, 
and  transmitter  loading  ca- 
pacitor. 

Rows  of  holes  are  drilled  in 
the  transmitter  chassis  to  im- 
prove ventilation  of  the  rectfier 
area  under  the  chassis. 


Figure  40 

UNDER-CHASSIS  VIEW,  R-F  DECKS 
OF  TRANSMITTER-RECEIVER 

Smo//  components  of  both  sections  are  mount- 
ed between  tube  socket  pins  and  phenolic 
terminal  strips  mounted  at  center  of  chassis. 
Smalt  copper  shields  are  placed  across  center 
of  6AJ4  r-f  tube  sockets  of  the  receiver. 
Shield  plate  made  of  perforated  aluminum 
sheet  is  placed  over  bottom  of  transmitter 
chassis  to  reduce  injection  voltage  to  receiver 
mixer  stage.  Neutraliiation  capacitors  of 
transmitter  amplifier  stage  are  mounted  di- 
rectly across  socket  on  either  side  of  the 
grid  coil. 

vides  the  proper  mixing  voltage  for  this  stage. 
Three  stages  of  6BA6  amplifiers  at  4.5  Me. 
comprise  the  i-f  section  of  the  receiver.  A-v-c 
is  applied  to  each  stage.  A 6T8  multi-purpose 
tube  (Vi)  is  employed  as  diode  detector,  a-v-c 
rectifier,  noise  limiter,  and  first  audio  stage. 

The  transmitter  section  is  mounted  upon  a 
separate  chassis.  A 6AB4  triode  oscillator  (Vs) 
operates  on  24-994  Me.  The  plate  circuit  of 
this  stage  is  split  to  provide  balanced  drive 
to  a push-pull  12AT7  triplet,  which  in  turn 


Figure  41 

TRANSMITTER.RECEIVER  CABINET, 
SHOWING  POWER  SUPPLY 
AND  MODULATOR 

The  power  supply  and  modulator  chassis  are 
mounted  vertically  against  the  side  walls  of 
the  cabinet.  The  tubes  are  mounted  on  the 
top  edge  of  the  chassis.  A small  blower  motor 
is  placed  at  the  rear  of  the  chassis  for  con- 
tinuous duty  operation.  R-f  section  sits  upon 
power  transformer  and  is  held  in  position  by 
panel  bolts  and  sheet  metal  screws  passed 
through  rear  of  cabinet. 


drives  a second  12AT7  tripler  stage  to  224.6 
Me.  All  stages  employ  capacity  coupling.  A 
third  12AT7  (Vn)  is  used  as  a push-pull  neu- 
tralized amplifier  at  224.6  Me.  The  plate  cir- 
cuit of  this  stage  is  a "hair  pin”  inductance 
(Ls)  tuned  by  a miniature  butterfly  capacitor 
C«.  Inductive  coupling  is  employed  between  the 
driver  stage  (Vio)  and  the  amplifier,  and  be- 
tween the  amplifier  and  the  antenna  circuit.  A 
1N34  germanium  diode  serves  as  a r-f  volt- 
meter in  the  antenna  circuit  for  transmitter 
tuning  purposes. 

The  transmitter  modulator,  audio  tone  os- 
cillator, and  receiver  audio  amplifier  stage  are 
mounted  upon  another  chassis  (figure  38). 
The  modulator  uses  two  tubes.  A 12AU7  (Vu) 
serves  as  a two  stage  resistance  coupled  speech 
amplifier,  and  a 12AX7  (Via)  is  used  as  a 
class  B modulator.  The  speech  amplifier  is  de- 
signed to  be  used  with  a mobile-type  carbon 
microphone.  M-c-w  transmission  is  permissible 
in  the  220  Me.  band  and  a 12AU7  (Via)  audio 
oscillator  has  been  added  for  this  purpose.  The 
oscillator  may  be  keyed  for  code  transmissions. 

The  power  supply  occupies  the  remaining 
chassis.  It  uses  two  5Y3  rectifier  tubes  to  pro- 
vide 300  volts  at  a current  drain  of  210  mil- 
liamperes.  No  standby  circuits  are  required 
since  both  transmitter  and  receiver  operate 
simultaneously. 

Assembly  The  complete  assembly  is  hous- 
and  Wiring  ed  in  a steel  cabinet  measuring 
1"  X 12"  X 8"  (figures  34  and 
36).  The  receiver  and  transmitter  portions  are 
built  upon  two  aluminum  chassis  measuring 
5"  X 8"  X 1".  These  chassis  are  bolted  together 
as  seen  in  figures  39  and  40  and  occupy  the 
center  section  of  the  cabinet.  The  power  supply 
and  modulator  are  built  upon  two  similar 


602  Low  Power  Transmitters 


THE  RADIO 


chassis  that  are  mounted  against  the  side  walls 
of  the  cabinet  as  seen  in  figure  41.  The  corners 
of  these  chassis  are  rounded  to  permit  mount- 
ing them  snugly  against  the  walls  of  the  cab- 
inet. The  various  tubes  are  mounted  in  a verti- 
cal position  on  the  upper  side  of  the  chassis 
while  the  transformers  mount  on  the  vertical 
surface.  To  the  rear  of  the  cabinet  is  a small 
blower  motor  to  provide  adequate  ventilation 
for  protracted  operation  of  the  equipment.  The 
various  manual  controls  fasten  directly  to  the 
main  panel  of  the  cabinet  and  flexible  leads 
are  run  from  them  to  their  respective  circuits 
in  the  equipment.  The  r-f  deck  sits  atop  the 
power  transformer  and  the  modulation  trans- 
former and  is  held  in  position  by  sheet  metal 
screws  that  pass  through  the  rear  wall  of  the 
cabinet  into  the  chassis. 

The  Transmitter  Chassis-  The  placement  of 
major  parts  on  the  transmitter  chassis  may  be 
seen  in  figures  39  and  40.  The  crystal  oscilla- 
tor stage  is  at  the  front  of  the  chassis  with  the 
crystal  mounted  in  a horizontal  position,  pro- 
jecting through  the  front  panel  of  the  cabinet. 
The  multiplier  stages  fall  in  a line,  with  the 
final  amplifier  stage  at  the  rear  of  the  chassis. 
The  coaxial  antenna  receptacle  Ji  and  antenna 
tuning  capacitor  C?  are  mounted  to  the  rear, 
wall  of  the  chassis. 

Common  v.h.f.  wiring  techniques  are  em- 
ployed in  the  assembly.  Short,  direct  leads  in 
the  r-f  section  are  mandatory.  The  sockets  of 
the  two  triplet  tubes  (V»  and  Vio)  and  the 
amplifier  tube  (Vn)  ate  positioned  at  an 
angle  of  45  degrees  so  that  a line  drawn 
through  pins  1 and  6 is  parallel  with  the  front 
edge  of  the  chassis.  The  midget  butterfly  ca- 
pacitors are  mounted  in  line  with  the  tube 
sockets,  permitting  very  short  leads  to  be  run 
from  the  plate  pins  of  the  tubes  to  the  stators 
of  the  capacitors. 

The  nine-pin  tube  sockets  are  the  type  that 
mount  from  beneath  the  chassis  and  have  a 
metal  ring  encasing  the  phenolic  portion  of  the 
socket.  The  various  pins  of  each  socket  that 
must  be  grounded  are  bent  down  and  soldered 
directly  to  this  ring.  The  grid  resistors  of  Vo 
and  Vio  are  installed  directly  between  the  grid 
pins  and  the  grounded  cathode  pins  of  the  tube 
socket.  The  coupling  capacitors  between  the 
stages  are  placed  between  the  grid  pins  and 
the  stator  rods  of  the  butterfly  tuning  capaci- 
tors. Filament  pins  4 and  5 of  the  nine-pin 
sockets  are  grounded  to  the  socket  ring  and  a 
common  filament  lead  runs  between  the  num- 
ber 9 pins. 

Inductors  Li  and  L2  are  made  of  manufac- 


tured coil  Stock  and  are  mounted  to  the  stator 
rods  of  the  tuning  capacitors.  Coils  Ls  and  L4 
may  be  wound  from  enameled  wire.  The  var- 
ious power  leads,  decoupling  resistors  and  ca- 
pacitors are  placed  clear  of  the  r-f  circuitry 
and  the  components  are  mounted  upon  pheno- 
lic tie  point  strips  placed  adjacent  to  the  tube 
sockets.  The  power  leads  and  meter  leads 
that  leave  the  chassis  pass  through  ceramic 
feed  through  capacitors  (Centralab  FT -1000) 
mounted  on  the  walls  of  the  chassis. 

Preliminary  Transmitter  Adjustments.  Trans- 
mitter operation  may  be  checked  using  the 
power  supply  shown  in  figure  38,  or  any  aux- 
iliary supply  capable  of  providing  about  200 
volts  at  80  milliamperes.  If  the  permanent  sup- 
ply is  used  for  tune-up  purposes,  a 5000  ohm, 
10-watt  dropping  resistor  should  be  placed  in 
series  with  the  B-plus  lead  to  the  exciter  to 
protect  the  tubes  in  case  of  misadjustments. 
Tuned  circuit  Li-G  may  be  set  to  25  Me.  with 
the  aid  of  a grid  dip  oscillator  and  the  6AB4 
oscillator  tube  and  crystal  Xi  plugged  in  their 
sockets.  Proper  operation  of  this  and  succeeding 
stages  may  be  checked  with  a 2 volt,  60  milli- 
ampere  (pink  bead)  flash  light  bulb  attached 
to  a small  loop  of  wire  which  is  held  in  prox- 
imity to  the  coil  of  the  stage  being  adjusted. 
When  oscillation  is  obtained,  the  12AT7  trip- 
let tube  is  plugged  in  its  socket  and  circuit 
Li-Ci  set  to  75  Me.  with  the  grid  dipper.  Plate 
voltage  is  again  applied  and  the  circuits  touch- 
ed up  for  maximum  bulb  brilliance  when  the 
pickup  loop  is  held  near  1=.  The  second  triplet 
tube  Vio  is  plugged  in  the  socket  and  circuit 
La-Cs  adjusted  for  maximum  output  at  225 
Me.  using  the  indicator  lamp.  Spacing  of  the 
turns  of  L.1  may  require  adjustment  to  permit 
resonance.  Plate  current  drawn  by  these  three 
tubes  should  be  about  70  milliamperes  when 
the  dropping  resistor  is  shorted  out. 

The  next  step  is  to  neutralize  the  amplifier 
stage.  This  may  be  done  by  observing  the  grid 
current  of  Vn  while  capacitors  C,  and  Cs  are 
adjusted.  The  leads  of  a high  resistance  volt- 
meter are  temporarily  clipped  across  the  2.7K 
grid  resistor  of  the  amplifier  stage,  and  the 
exciter  circuits  are  repeaked  for  maximum  grid 
voltage  reading.  Over  twenty  volts  should  be 
obtained.  Plate  voltage  to  the  amplifier  stage  is 
removed  for  this  test.  Neutralizing  capacitors 
Cl  and  G are  now  set  to  minimum  capacity 
and  coil  Li  is  varied  in  position  with  respect 
to  Li  to  obtain  maximum  grid  voltage.  The 
amplifier  plate  tuning  capacitor  Co  is  tuned 
through  resonance  while  grid  voltage  is  ob- 
served. If  a change  in  voltage  occurs  during 


HANDBOOK 


220-Mc.  Duplex  603 


the  tuning  process  the  settings  of  capacitors 
C4  and  G are  advanced  in  unison  until  no 
change  of  grid  voltage  is  noticeable.  As  the 
capacity  settings  are  increased  (keeping  the 
two  settings  approximately  equal)  the  "kick” 
of  grid  voltage  will  gradually  decrease,  until 
there  is  either  no  movement  of  the  meter  as 
G is  varied,  or  else  merely  a gradual  rise  of 
a volt  or  so  as  Cn  is  swung  through  its  range. 

A 6.3  volt,  150  milliampere  (brown  bead) 
lamp  is  now  connected  to  antenna  receptacle 
Ji  and  plate  voltage  is  applied  to  the  amplifier 
stage  and  exciter.  Capacitors  Cc  and  Ct  are 
tuned  for  maximum  bulb  brilliance.  The  plate 
current  of  the  amplifier  stage  will  be  close  to 
30  milliamperes  under  conditions  of  maximum 
output. 

Once  initial  adjustments  have  been  made, 
a 0-1  d.c.  milliammeter  may  be  connected  to 
the  r-f  voltmeter  terminals  and  the  transmitter 
may  be  completely  tuned  by  adjusting  all  re- 
sonant circuits  for  maximum  indication  of  the 
meter.  It  is  wise,  however,  to  make  the  pre- 
liminary adjustments  stage  by  stage  as  des- 
cribed above  to  acquaint  the  operator  with 
proper  transmitter  operation. 

The  Receiver  Chassis-  The  placement  of  the 
major  components  of  the  receiver  chassis  may 
be  seen  in  figures  39  and  40.  The  line-up  of 
stages  forms  a "U”  on  the  chassis,  passing 
down  one  side,  across  the  front  and  back  along 
the  other  side.  Two  six  terminal  phenolic  tie- 
point  strips  ate  mounted  along  the  center  line 
of  the  chassis,  supporting  various  decoupling 
and  voltage  dropping  resistors.  Layout  of  the 
r-f  stages  may  be  seen  in  the  under-chassis 
view  of  figure  40.  A small  copper  shield  plate 
1"  high  and  1 Vi  inches  long  passes  across  the 
center  of  each  socket  and  is  grounded  at  each 
end  by  soldering  to  the  socket  retaining  screws. 
The  center  stud  of  the  socket,  and  pins  1,  3, 
4,  6,  and  9 are  all  soldered  to  this  plate.  Tun- 
ing capacitors  G,  G,  Cw,  and  Cn  are  mounted 
close  to  the  tube  sockets,  and  the  VHF-type 
bypass  capacitor  at  the  "cold”  end  of  each  coil 
(Centralab  type  ZA)  is  mounted  next  to  the 
tuning  capacitor.  Coils  Ls,  Lo,  Lio,  and  Lh  are 
mounted  between  the  terminals  of  the  respec- 
tive tuning  capacitor  and  the  adjacent  VHF 
bypass  capacitor.  Care  must  be  taken  to  keep 
the  coil  leads  as  short  as  possible — less  than 
VV-inch  long  in  any  case. 

One  filament  choke  of  each  r-f  amplifier 
stage  is  grounded  to  the  shield  plate  and  the 
Other  choke  is  bypassed  by  a disc  ceramic  ca- 
pacitor and  returns  to  the  common  filament 
lead.  Wiring  of  the  i-f  stages  is  straightforward, 


VERTICAL  ARRAY  HORIZONTAL  ARRAY 

FOR  TRANSMISSION  FOR  RECEPTION 


ELEMENT  LENGTHS 

AND  SPACING 

REFLECTOR(R) 

25,  5" 

ANTENNA  fA) 

23,5" 

DIRECTOR  (D) 

22.25' 

SPACING  (S) 

10" 

COAXIAL  BALUN 
TOTAL  LENCTH  * 1 7 1 /4 
MADE  OF  79  Ji  COAXIAL  LINi 


75  fl  COAXIAL  LINE 
TO  TRANSMITTER- RECEIVER 


1^300  n.  TV  LINE 
{random 

lencth) 


^ SOLDE  R 3 
SHIELDS 


Figure  42 

DUAL  ANTENNA  SYSTEM  FOR  220 
MC.  TRANSMITTER-RECEIVER 

The  same  polariiaiion  must  be  used  on  each 
link.  This  means  that  polarization  of  trans- 
mission and  reception  is  different  at  each  sta- 
tion. For  the  companion  link  to  the  one  illu- 
strated in  this  drawing,  the  vertical  array 
should  be  used  for  reception,  and  the  hori- 
zontal array  for  transmission. 


components  being  mounted  between  the  socket 
and  i-f  transformer  terminals  and  the  phenolic 
tie-point  strip.  Power  leads  are  terminated  at 
the  six  prong  plug  mounted  on  the  rear  apron 
of  the  chassis. 

No  connection  need  be  made  between  the 
r-f  circuit  of  the  transmitter  and  the  mixer 
stage  of  the  receiver  as  ample  proximity  coupl- 
ing exists.  In  fact,  a shield  plate  is  placed  over 
the  bottom  of  the  transmitter  chassis  to  reduce 
the  mixer  injecion  to  a tolerable  level. 


Preliminary  Receiver  Adjustment.  After  all 
wiring  is  completed,  the  i-f  section  of  the  re- 
ceiver may  be  aligned.  A 4.5  Me.  tone  modu- 
lated signal  is  loosely  coupled  to  the  plate  pin 
( # 5 ) of  mixer  tube  Va  and  earphones  or  an 
a-c  meter  connected  to  the  output  circuit  of 
audio  tube  V-  (Pins  5 and  6 of  the  power 
plug).  The  slug  cores  of  transformers  Ti,  Ti, 
Ta,  and  Ti  are  adjusted  for  maximum  signal. 
The  receiver  and  transmitter  sections  should 
be  bolted  together  and  the  transmitter  tuned 


Figure  43 

COMPACT  VFO  PROVIDES  HIGH 
STABILITY  SIGNAL  FOR  HIGH 
FREQUENCY  DX  BANDS 

This  small  v.f,o.  employs  latest  techniques  to 
achieve  stable^  drift-free  signal.  Only  7V2’'  X 
9**  X 7"  in  size,  the  unit  uses  a high-C  7.75 
Me.  oscillator  followed  by  a cathode  follower 
to  achieve  a maximum  degree  of  circuit  isola- 
tion. Buffer  tuning  and  control  switch  are  at 
left  of  panel,  with  large,  easily  read  fre- 
quency control  dial  at  center. 


Up  wih  a dummy  load.  This  will  provide  in- 
jecion  signal  for  proper  receiver  operation  at 
220.1  Me.  A remote  signal  (preferably  that 
of  the  companion  unit)  should  be  used  as  a 
signal  source  and  the  tuned  circuits  of  the  r-f 
amplifiers  adjusted  for  best  signal  reception. 
It  may  be  necessary  to  alter  the  spacing  of  coils 
Ls  - Lii  to  provide  proper  circuit  resonance 
which  should  occur  at  about  the  middle  of  the 
tuning  range  of  the  associated  capacitor. 

Modulator  and  Power  Supply  Construction. 
The  layout  of  the  modulator  and  power  supply 
can  be  seen  in  figure  41.  The  units  are  wired 
and  the  control  leads  are  brought  out  through 
insulated  grommets  placed  in  the  end  of  the 
chassis.  A common  ground  lead  is  run  from 
each  chassis  to  the  switches  and  controls  mount- 
ed upon  the  front  panel.  The  small  blower 
motor  is  mounted  at  the  rear  of  the  cabinet 
and  a row  of  ventilation  holes  is  drilled 
around  the  top  edge  of  the  cabinet  as  shown 
in  the  photograph.  The  power  supply  and 
modulator  should  be  tested  before  they  are 
mounted  in  the  cabinet,  and  the  whole  assem- 
bly should  be  in  operating  condition  before 
it  is  placed  within  the  cabinet.  The  tone  of 
the  audio  oscillator  (V15)  may  be  varied  by 
placing  a capacitor  across  the  primary  winding 
of  the  oscillator  transformer  Ts. 

Antenna  Two  separate  antennas  are  re- 

Installation  quired  for  duplex  operation.  To 
reduce  coupling,  one  link  is 
hori2ontally  polari2ed  and  the  other  link  is 
vertically  polarized  as  illustrated  in  figure  42. 
Care  must  be  taken  that  identical  polarization 
is  employed  at  both  ends  of  each  link.  Three- 
element  beam  antennas  are  used  for  each  link. 
A folded  dipole  and  balun  system  are  employed 
to  provide  a good  match  to  the  coaxial  trans- 
mission line.  After  the  links  have  been  set  up, 
transmitter  and  receiver  tuning  adjustments 
should  be  peaked  to  maximize  the  signals.  It 
will  be  found  advantageous  to  experiment  with 


the  position  of  the  antennas  to  provide  max- 
imum signal  pickup  with  a minimum  of  inter- 
ference from  the  local  transmitter.  In  general, 
the  receiver  antenna  should  be  oriented  for 
maximum  signal  from  the  distant  station,  and 
the  transmitter  antenna  should  be  oriented  for 
minimum  local  signal  pickup.  For  general  use, 
the  two  antennas  may  be  attached  to  the  same 
supporting  structure,  separated  from  each  other 
by  four  or  five  feet.  As  with  any  VHF  equip- 
ment, time  spent  in  placing  the  antennas  will 
pay  big  dividends  in  system  operation. 

28-6  A High  Stability 
V.F.O.  For  the 
DX  Operator 

Stability  and  freedom  from  drift  are  the 
prime  requisites  of  a high  quality  variable  fre- 
quency oscillator.  To  meet  the  requirements 
of  the  discriminating  operator  the  v.f.o.  must 
be  stable  with  respect  to  warm-up,  ambient 
temperature  variations,  line  voltage  shift,  vi- 
bration, and  the  presence  of  a strong  r-f  field. 
To  achieve  these  parameters  in  a unit  simple 
enough  for  home  construction  is  a large  order. 

The  variable  frequency  oscillator  shown  in 
figures  43,  46,  and  47  is  a successful  attempt 
to  meet  these  requirements  and  is  recommend- 
ed to  those  operators  who  desire  a v.f.o.  having 
a high  order  of  stability  and  resetability. 

Frequency  The  perfect  variable  frequency 

Stability  oscillator  has  the  frequency  de- 

termining elements  completely 


VOLTAGE  MEASUREMENTS, 

V.  F.O. 

SCREEN  {PIN  6Ue— 70 

CATHODE  (P/N8)  6U8—  10.5 

R-F  VOLTS,  GRID  (P/N9)  6U8—  10 

3 

R-F  VOLTS,  CATHODE  (P/NS)  6U8- 

- 9.5 

R-F  VOLTS,  Jl.  ACROSS  47  OHM  RES 

ISTOR  -6.2 

Figure  44 

VOLTAGE  CHART  FOR  V.F.O. 


HANDBOOK 


High-Stability  V.F.O.  605 


C]— .002  fifd.  Ceniratab  high  accuracy  capacitor  950-202. 

C2_300  fip-fd.  Bud  MC-1860 
Cj_25  fip-fd.  Bud  LC-T642 

turns  #20  e,  3/^"  diam.,  approx.  I"  long.  (See  text)  Wind  on  National  ceramic  form  XR-72.  Top 
ot  4 turns  from  "cold"  end. 

Li— 4.5  Me.  Interstage  "TV"  transformer.  J.  W.  Miller  T466.  Remove  secondary  winding  and  replace 
with  8 turns  #22  d.c.c.  wound  next  to  primory  winding. 

PFCi— 2.5  mh.  National  R-IOO 
RFCt-—VHF  choke,  500  mo.  Notionof  R-60 


isolated  both  electrically  and  physically  from 
the  test  of  the  transmitting  equipment.  This 
goal  cannot  be  achieved  in  practice  since  a 
vacuum  tube  ot  transistor  must  be  coupled  to 
these  circuit  elements  to  maintain  oscillation. 
The  coupling  of  such  items  deteriorates  the 
electrical  isolation  of  the  frequency  determin- 
ing circuit.  By  the  use  of  proper  coupling  cir- 
cuits the  effects  of  the  tube  ot  transistor  can 
be  minimized. 

The  frequency  determining  circuit  is  also. 


subject  to  temperature  variations,  changes  in 
the  relative  humidity  of  the  surrounding  air, 
and  absorption  of  heat  from  nearby  objects.  By 
employing  temperature  and  humidity  resistive 
materials  and  removing  as  many  heat  generat- 
ing components  from  the  vicinity  of  the  fre- 
quency determining  circuits,  a good  degree  of 
stability  in  the  v.f.o.  may  be  obtained. 

Finally,  steps  must  be  taken  to  ensure  that 
the  oscillator  of  the  v.f.o.  unit  is  completely 
free  of  parasitics,  and  that  the  output  waveform 


Figure  46 

REAR  VIEW  OF  V.F.O.  CHASSIS, 
SHOWING  PLACEMENT  OF  PARTS 

Oscillator  inductor  Li  is  at  the  left  of  chassis, 
separated  from  the  heat  producing  electron 
tubes.  To  right  of  L,  is  the  precision  padding 
capacitor  Cr  mounted  to  the  chassis.  The  main 
tuning  capacitor  is  firmly  affixed  to  a heavy 
dural  mounting  plate  bolted  to  the  gear  box 
of  the  dial  drive.  The  rear  of  the  tuning  ca- 
pacitor is  attached  to  the  chassis  by  a mount- 
ing stud.  To  the  right  are  the  three  tubes  and 
the  output  transformer.  Directly  under  the 
tuning  capacitor  is  the  polystyrene  chassis 
plate  holding  the  feed'  through  bushings  for 
the  oscillator  leads  to  the  under-chassis  area. 

Control  switch  Si  is  at  upper  right. 


606  Low  Power  Transmitters 


THE  RADIO 


Figure  47 

UNDER-CHASSIS  VIEW  OF 
V.F.O.  UNIT 

Power  lead  filters  are  at  lower  right  of  chas- 
sis, with  r-f  circuit  components  at  left.  All 
parts  are  firmly  mounted  to  terminals  and 
tie  points  to  reduce  vibration.  Power  leads 
are  laced.  Buffer  tuning  capacitor  Cj  is  at 
upper  left. 

is  low  in  harmonic  content.  The  variable  fre- 
quency exciter  to  be  described  in  this  section 
meets  these  fundamental  requirements. 

The  V.F.O.  Circuit  The  circuit  diagram  of 
the  high  stability  v.f.o. 
is  shown  in  figure  45,  and  the  complete  physi- 
cal layout  may  be  seen  in  figures  46  and  47. 

The  pentode  section  of  a miniature  6U8 
tube  is  used  as  the  oscillator.  The  oscillator 
circuit  is  tuned  to  the  160  meter  region  and 
consists  of  a high-Q  oscillator  coil  resonated 
to  the  working  frequency  by  a precision  ceram- 
ic capacitor  in  parallel  with  a variable  air 
capacitor.  The  ceramic  capacitor  ( Centrdab 
type  950)  has  a measured  temperature  coeffi- 
cient of  less  than  plus  or  minus  ten  parts  per 
million  over  the  temperature  range  of  — 40 
degrees  to  plus  60  degrees  Centigrade.  The 
use  of  ordinary  ceramic  or  silver  mica  capaci- 
tors as  a substitute  for  this  unit  is  not  recom- 
mended, since  the  temperature  coefficient  of 
such  units  cannot  be  closely  controlled  in 
quantity  production. 

The  oscillator  coil,  Li,  is  wound  upon  a cer- 
amic coil  form  that  has  a very  low  coefficient 
of  expansion.  The  coil  turns  are  spaced  to  en- 
sure low  inter-turn  capacitance.  The  wire  is 
wound  upon  the  form  under  tension  and  at  a 
relatively  high  temperature.  As  the  wire  cools, 
it  shrinks  and  tightly  clasps  the  form  so  that 
for  all  practical  purposes  the  wire  and  the 
form  have  the  same  coefficient  of  expansion 


with  regard  to  temperature  changes.  This  style 

of  winding  reduces  the  possibility  of  the  oscil- 
lator developing  random  frequency  jumps  with 
changes  in  temperature  and  humidity. 

A 56  ohm  resistor  is  inserted  in  series  with 
the  grid  lead  of  the  6U8  oscillator  tube.  This 
suppressor  eliminates  any  tendency  toward  par- 
asitic oscillation  that  can  result  in  this  type 
of  oscillator  circuit.  The  parastic  tends  to  make 
fundamenal  oscillation  unstable  at  certain  set- 
tings of  the  oscillator  tuning  capacitor.  This 
instability  may  be  caused  by  variations  of  the 
inherent  inductance  of  the  tuning  capacitor  at 
the  frequency  of  parasitic  oscillation. 

All  components  of  the  oscillator  circuit  are 
spaced  clear  of  the  oscillator  tube  to  prevent 
heat  transfer  from  the  tube  to  the  frequency 
determining  circuit. 

The  pentode  oscillator  is  R-C  coupled  to 
the  triode  section  of  the  6U8  which  serves  as 
a cathode  follower.  The  input  impedance  of 
the  cathode  follower  is  extremely  high  and 
provides  excellent  isolation  between  the  oscil- 
lator circuit  and  the  output  stage.  Plate  and 
screen  voltage  of  the  oscillator  and  plate  volt- 
age of  the  cathode  follower  are  regulated  by  an 
OA2  gas  regulator  to  provide  maximum  iso- 
lation from  power  supply  fluctuations. 

The  r-f  output  of  the  triode  section  of  the 
6U8  is  taken  from  the  low  impedance  cathode 
circuit  and  is  capacity  coupled  to  a 6AQ5  min- 
iature pentode  tube  which  serves  as  a frequency 
doubler  to  the  80  meter  region.  The  v.f.o. 
covers  the  frequency  range  of  1750  kc.  to 
1850  kc.,  and  the  plate  circuit  of  the  6AQ5 
doubler  tunes  3500  kc.  to  3700  kc.  The  80 
meter  output  is  0.8  watt  which  is  more  than 
sufficient  to  drive  a tetrode  buffer  such  as  an 
807  or  6146.  If  more  output  is  desired,  a 
6CL6  may  be  substituted  for  the  6AQ5  (with 
appropriate  socket  wiring  changes)  to  deliver 
approximately  2 watts. 

One  filament  terminal  of  each  tube  is 
grounded  and  the  free  terminals  are  bypassed 
to  ground  with  .005  gfd.  ceramic  capacitors. 
The  filament  leads  then  pass  through  an  r-f 
choke  to  ensure  maximum  lead  isolation. 

Switch  Si  disables  the  v.f.o.  by  removing  the 
plate  voltage.  A second  section  of  the  switch 
opens  an  auxiliary  circuit  that  can  control  the 
power  relays  of  the  transmitter.  A third  posi- 
tion of  switch  Si  permits  the  v.f.o.  to  be 
turned  on  by  itself  for  zero-beat  operation. 

Transmitter  keying  for  c-w  operation  is  done 
in  the  stages  following  the  v.f.o.-exciter  to 
achieve  maximum  frequency  stability  and  free- 
dom from  frequency  shift  during  keying. 


HANDBOOK 


High-Stability  V.F.O.  607 


Mechanical  Desisn  The  mechanical  design 
of  the  V.F.O.  of  the  variable  frequen- 

cy oscillator  is  equally 
important  as  the  electrical  design  if  maximum 
stability  and  reliability  ate  to  be  achieved. 
Provision  must  be  made  for  dissipation  of  the 
heat  generated  by  the  vacuum  tubes,  and  the 
v.f.o.  components  should  be  mounted  on  a 
sheet  of  conducting  material  which  will  act  as 
a heat  "sink”  and  will  tend  to  resist  rapid 
changes  in  the  temperature  of  the  components 
bolted  thereto. 

The  v.f.o.  is  built  upon  a cadmium  plated 
steel  chassis  measuring  IV2"  x 9”  x 114".  The 
gear  box  of  the  National  HRO-type  dial  is 
Isolted  to  the  chassis  and  the  main  tuning  ca- 
pacitor G is  affixed  by  its  front  bearing  to  a 
3"  X 3”  X 0.125"  dural  plate  held  to  the  ca- 
pacitor gear  box  by  three  long  machine  screws 
and  three  I Vs"  metal  spacers.  The  tuning  ca- 
pacitor is  driven  through  a high  quality  flexi- 
ble coupling.  This  coupling  should  be  free  of 
back-lash,  and  should  not  permit  end  pressure 
on  the  shaft  of  the  capacitor.  A Johnson  104- 
250  coupling  is  recommended.  Make  sure  that 
the  shaft  of  the  gear  box  and  that  of  the  ca- 
pacitor are  perfectly  in  line  as  misalignment 
tends  to  force  a degree  of  back-lash  in  the 
system. 

The  rear  mounting  foot  of  the  capacitor  is 
fastened  to  the  steel  chassis  by  means  of  a long 
6/32  bolt  and  a 1-inch  metal  spacer.  This  as- 
sembly provides  a very  rugged  arrangement 
with  a minimum  of  flexing  between  the  ca- 
pacitor and  the  chassis. 

Oscillator  coil  Li  and  the  precision  ceramic 
capacitor  Ci  are  mounted  to  the  chassis  to  the 
side  of  the  tuning  capacitor  as  seen  in  the  rear 
view  photograph.  The  slug  of  coil  Li  is  re- 
moved and  discarded  before  the  coil  is  wound. 
Moveable  oscillator  slugs  are  not  conducive 
to  any  great  degree  of  oscillator  stability  and 
this  potential  source  of  instability  should  be 
taken  from  the  circuit  before  it  can  do  any 
damage. 

The  leads  from  the  oscillator  circuit  to  the 
6U8  pentode  tube  pass  through  a 114"  hole 
cut  in  the  steel  chassis.  A Vs  ’-thick  square  of 
polystyrene  is  mounted  under  this  hole  and  two 
miniature  feedthrough  bushings  are  mounted 
in  the  insulating  plate.  The  residual  capacity 
to  grounds  of  these  important  leads  is  thereby 
held  to  an  absolute  minimum  value. 

The  vacuum  tubes  and  auxiliary  circuits 
are  placed  on  the  opposite  side  of  the  tuning 
capacitor  from  the  sensitive  tuned  circuit  Li-G. 
In  front  of  the  oscillator  tube  is  the  80  meter 


output  coil  L2,  the  6AQ5  socket  and  the  OA2 
socket.  Switch  Si  is  mounted  to  the  front  panel 
directly  in  front  of  the  6AQ5.  The  front  panel 
is  held  to  the  chassis  by  means  of  four  6/32 
bolts  placed  in  the  extreme  corners  of  the 
front  lip  of  the  chassis.  The  panel  is  spaced 
slightly  away  from  the  chassis  by  an  extra  set 
of  nuts  slipped  over  the  bolts  before  the  panel 
is  affixed  to  the  chassis. 

All  control  leads  to  the  v.f.o.  unit  pass 
through  a four  wire  cable  which  enters  the  un- 
der chassis  area  via  a rubber  grommet  on  the 
back  lip  of  the  chassis.  Each  lead  is  bypassed 
with  a .005  /ifd.  ceramic  capacitor  and  is  fil- 
tered with  a low  resistance  r-f  choke.  The  co- 
axial output  receptacle  for  the  v.f.o.  is  also 
mounted  on  the  tear  lip  of  the  chassis,  directly 
behind  the  6U8  oscillator  tube. 

Wiring  the  V.F.O.  The  under-chassis  area  of 
the  v.f.o.  should  be  wired 
first.  Common  ground  connections  are  made 
to  the  three  sockets  by  means  of  soldering  lugs 
and  lock  washers  placed  beneath  one  socket  re- 
taining nut.  The  filament  leads  are  wired  next. 
The  bypass  capacitors  for  pins  1,  3,  and  4 of 
the  6U8  socket  are  placed  between  the  socket 
pins  and  the  grounded  center  stud  of  the  socket 
with  the  shortest  possible  leads.  All  compon- 
ents of  the  6U8  stage  should  be  mounted  solid- 
ly in  place  between  adjacent  tube  pins  or  to 
nearby  phenolic  tie-point  terminals.  The  grid 
resistor  and  capacitor  of  the  oscillator  section 
are  mounted  between  pin  #2  of  the  6U8 
socket  and  the  inter-chassis  feedthrough  insu- 
lator in  the  polystyrene  block. 

The  plate  coil  of  the  6AQ5  doubler  stage  is 
made  from  a 4.5  Me.  sound  TV  i-f  transformer. 
The  secondary  winding  and  tuning  capacitor 
are  removed  and  a link  winding  of  8 turns  is 
wound  on  the  form,  closely  spaced  to  the  pri- 
mary winding.  The  transformer  is  then  re- 
placed in  the  shield  can  and  mounted  atop  the 
chassis.  The  r-f  choke  in  the  plate  circuit  of 
the  doubler  stage  is  mounted  between  two  ter- 
minals of  a phenolic  tie-point  strip  bolted  to 
one  of  the  transformer  lugs. 

The  power  supply  r-f  filter  circuits  are 
placed  on  two  four  terminal  tie-point  strips 
mounted  in  the  far  corner  of  the  chassis.  The 
bypass  capacitors  are  mounted  between  the  in- 
dividual terminal  lugs  and  the  adjacent  ground 
connections  of  the  strip  and  the  r-f  chokes  are 
mounted  between  the  strips. 

After  the  under-chassis  wiring  is  completed 
the  wiring  atop  the  chassis  may  be  done.  The 
leads  of  the  tuned  circuit  ate  made  of  #12 
tinned  wire.  The  coil  is  wound  with  #20 


608  Low  Power  Transmitters 


THE  RADIO 


enameled  wire  space-wound  to  a length  of  ap- 
proximately one  inch.  The  easiest  way  to  ob- 
tain the  correct  spacing  is  to  wind  two  separate 
windings  on  the  coil.  One  winding  is  the  de- 
sired one  of  #20  wire,  the  second  is  composed 
of  #24  enameled  wire  and  is  used  only  for 
spacing.  The  two  windings  are  put  on  simul- 
taneously and  the  #24  wire  is  removed  after 
the  other  winding  has  been  properly  termin- 
ated at  both  ends.  To  obtain  the  best  tempera- 
ture coefficient  of  the  coil  the  wire  should  be 
wound  on  when  it  is  warm.  The  easiest  way  to 
accomplish  this  is  to  place  rolls  of  the  two 
wire  sizes  in  the  open  and  warm  them  slowly 
until  they  are  almost  too  hot  to  touch.  The 
wires  are  then  removed  from  the  oven  and 
wound  on  the  coil  form  before  they  have  a 
chance  to  cool  to  room  temperature.  The  wire 
is  kept  under  tension  during  the  winding  pro- 
cess and  may  be  reheated  with  a soldering  iron 
or  heat  lamp  if  it  begins  to  grow  cold.  When 
the  coil  is  completed  and  the  wire  cools,  it 
will  be  found  to  be  tightly  wound  around  the 
coil. 

After  the  coil  has  been  completed  the  fourth 
turn  (cathode  tap)  should  be  cleaned  care- 
fully with  a small,  sharp  knife  blade  and  a 
short  length  of  #22  wire  is  soldered  to  the 
turn.  The  opposite  end  of  this  lead  is  attached 
to  a lug  terminal  at  the  base  of  the  coil.  All 
the  unused  lugs  may  then  be  temoved  from  the 
coil  form.  A separate  ground  lead  is  run  from 
the  ground  lug  of  the  coil  form  to  the  ground 
lug  of  Cl  and  then  to  the  rotor  terminal  of 
mning  capacitor  Ca  to  prevent  any  intermittent 
ground  paths  through  the  chassis  and  the  ca- 
pacitor mounting  assembly. 

V.F.O.  Power  The  schematic  of  the  v.f.o. 

Supply  power  supply  is  shown  in 

figure  48.  The  power  supply 
is  constructed  as  a separate  unit  since  it  is  de- 
sired to  keep  the  heat  and  vibration  of  the 
transformers  and  chokes  away  from  the  v.f.o. 
circuits.  The  power  construction  is  straightfor- 
watd,  with  all  leads  bypassed  to  prevent  r-f 
pickup  from  the  transmitter.  The  supply  is 
built  upon  a IVi"  x 5 Vi"  miniature  amplifier 
foundation  chassis  and  provides  300  volts  at 
50  milliamperes  for  the  v.f.o.  circuits. 

V.F.O.  Alignment  After  the  v.f.o.  wiring 

and  Adjustment  has  been  checked  the 

6U8  and  OA2  tubes 
should  be  plugged  in  their  sockets.  Switch  Si 
is  set  to  the  "off”  position  and  the  power  sup- 
ply is  turned  on.  The  filaments  of  the  tubes 
should  light,  and  when  Si  is  set  to  either  the 


T(  5V4-G 


nsv.'u 


Figure  48 

SCHEMATIC,  POWER  SUPPLY  FOR 
VFO 


Ti — 325*0*325  volts,  55  mg.  Staneor  PC*8407 
CHi,  CH2 — 7 henry  at  50  ma.  Staneor  C-1707 


"zero-beat”  or  "transmit”  position  the  v.f.o. 
signal  may  be  heard  in  a nearby  receiver  tuned 
to  the  160  meter  region.  The  oscillator  is  now 
run  for  a few  hours  before  it  is  adjusted  to 
the  correct  frequency  range.  The  next  step  is  to 
plug  the  6AQ5  in  its  socket,  and  a 6.3  volt, 
150  ma.  (brown  bead)  pilot  lamp  is  placed 
across  the  terminals  of  the  coaxial  output  jack 
Ji,  serving  as  a dummy  load.  The  v-f.o.  is 
turned  on  and  the  setting  of  Cs  and  the  slug 
of  coil  Ls  are  varied  to  provide  an  indication 
in  the  bulb. 

The  next  step  is  to  calibrate  the  oscillator.  A 
BC-221  frequency  meter  or  calibrated  receiver 
should  be  used.  The  dial  of  the  v.f.o.  should 
be  set  so  the  bandspread  scale  reads  "0”  when 
capacitor  C2  is  fully  meshed.  When  the  dial 
is  tuned  to  a reading  of  "10”  the  capacitor  is 
nearly  fully  meshed,  and  the  operating  fre- 
quency of  the  oscillator  should  be  3500.00  kc. 
Unless  you  are  extremely  lucky,  this  will  not 
be  the  case.  Since  there  are  no  adjustable  pad- 
ding capacitors  in  the  frequency  determining 
circuit  coil  Li  must  be  adjusted  a bit  at  a time 
until  the  correct  frequency  falls  at  the  desig- 
nated dial  reading.  The  turns  of  Li  may  be 
varied  in  position  with  the  aid  of  a pen-knife 
blade,  and  minute  adjustments  to  the  top  few 
turns  may  be  made  until  the  correct  calibra- 
tion is  reached-  Once  the  calibration  of  the 
oscillator  has  been  set,  the  turns  of  Li  may  be 
permanently  fixed  in  place  by  means  of  a few 
drops  of  colorless  nail  polish.  Use  the  mini- 
mum amount  of  polish  possible.  Placing  the 
polish  on  the  coil  will  vary  the  operating  fre- 
quency slightly,  so  a final  slight  adjustment 
must  be  made  after  the  last  spot  of  polish  has 
been  placed  on  the  coil.  Full  coverage  of  all 
the  amateur  bands  except  80  metets  will  be 
obtained  with  this  value  of  tuning  capacitor. 
Decreasing  the  size  of  C2  will  provide  greater 
bandspread  on  the  high  frequency  bands,  and 


HANDBOOK 


High-Stability  V.F.O.  609 


increasing  the  capacitance  of  G will  provide 
full  coverage  of  the  80  meter  band.  In  either 
case,  it  will  be  necessary  to  juggle  the  turns  on 
Li  to  obtain  the  correct  tuning  range. 

Operating  Check  When  completed  the  v.f.o. 
of  the  V.F.O.  should  be  placed  in  the 

operating  position  and  run 
for  a period  of  several  hours.  During  this  time, 
the  frequency  of  the  v.f.o.  should  be  compared 
against  a known  standard  such  as  a 100  kc. 
crystal,  or  Standard  Frequency  Station  WWV. 
Under  normal  conditions  the  v.f.o.  will  have 
a small  positive  initial  warm-up  drift  of  less 
than  100  cycles  on  80  meters.  It  should  settle 
down  after  a period  of  time  and  remain  rela- 
tively stable  if  the  temperature  of  the  room  is 
constant. 

Temperature  stabilization  may  be  accom- 
plished by  the  addition  of  a small  value  of 


negative  coefficient  capacitance  to  the  tuning 
circuit.  This  is  a cut-and-try  process  and  is  not 
recommended  unless  the  builder  has  plenty  of 
time  and  previous  experience  with  the  task. 
It  is  not  really  required  unless  the  oscillator  is 
operated  under  extremes  of  room  temperature. 

The  last  step  is  to  place  the  v.f.o.  in  the 
metal  cabinet.  The  unit  should  not  be  operated 
out  of  the  cabinet  as  the  enclosure  affords 
some  degree  of  shielding  from  the  t-f  field  of 
the  transmitter.  The  rear  of  the  chassis  is 
fastened  to  the  bottom  of  the  cabinet  by  two 
sheet  metal  screws  passed  from  the  under  side 
of  the  bottom  of  the  cabinet  up  into  the  rear 
lip  of  the  chassis.  To  prevent  frequency  changes 
during  operation  the  lid  of  the  cabinet  should 
be  fastened  shut  by  means  of  two  sheet  metal 
screws  run  through  the  front  corners  of  the 
lid  into  the  cabinet. 


CHAPTER  TWENTY-NINE 


The  trend  in  design  of  transmitters  for  oper- 
ation on  the  high  frequency  bands  is  toward 
the  use  of  a single  high-level  stage.  The  most 
common  and  most  flexible  arrangement  in- 
cludes a compact  bandswitching  exciter  unit, 
with  15  to  100  watts  output  on  all  the  high- 
frequency  bands,  followed  by  a single  power 
amplifier  stage.  In  many  cases  the  exciter  unit 
is  placed  upon  the  operating  table,  with  a co- 
axial cable  feeding  the  drive  to  the  power  am- 
plifier, although  some  operators  prefer  to  have 
the  exciter  unit  included  in  the  main  trans- 
mitter housing. 

This  trend  is  a natural  outgrowth  of  the  in- 
creasing importance  of  v-f-o  operation  on  the 
amateur  bands.  It  is  not  practical  to  make  a 
quick  change  in  the  operating  frequency  of  a 
transmitter  when  a whole  succession  of  stages 
must  be  returned  to  resonance  following  the 
frequency  change.  Another  significant  factor  in 
implementing  the  trend  has  been  the  wide  ac- 
ceptance of  commercially  produced  75  and 
150-watt  transmitters.  These  units  provide  r-f 


excitation  and  audio  driving  power  for  high- 
level  amplifiers  running  up  to  the  1000-watt 
power  limit.  The  amplifiers  shown  in  this 
chapter  may  be  easily  driven  by  such  exciters. 

29-1  Power  Amplifier 
Design 

Choice  of  Either  tetrode  or  triode  tubes  may 
Tubes  be  used  in  high-frequency  power 

amplifiers.  The  choice  is  usually 
dependent  upon  the  amount  of  driving  power 
that  is  available  for  the  power  amplifier.  If 
a transmitter-exciter  of  100-watt  power  capa- 
bility is  at  hand  (such  as  the  Heath  TX-1) 
it  would  be  wise  to  employ  a power  ampli- 
fier whose  grid  driving  requirements  fall  in 
the  same  range  as  the  output  power  of  the 
exciter.  Triode  tubes  running  1 -kilowatt  in- 
put (plate  modulated)  generally  require  some 
50  to  80  watts  of  grid  driving  power.  Such 
a requirement  is  easily  met  by  the  output 
level  of  the  100-watt  transmitter  which  should 


610 


Design  611 


TO  ANTENNA  TO  ANTENNA 

CIRCUIT  CIRCUIT 


UNBALANCED  COAXIAL  BALANCED  TWIN- LINE 

FEED  SYSTEM  FEED  SYSTEM 


Figure  1 

LINK  COUPLED  OUTPUT  CIRCUITS 
FOR  PUSH-PULL  AMPLIFIERS 

be  employed  as  the  exciter.  Tetrode  tubes 
(such  as  the  4-2 50 A)  require  only  10  to 
15  watts  of  actual  drive  from  the  exciter  for 
proper  operation  of  the  amplifier  stage  at  1- 
kilowatt  input.  This  means  that  the  output 
from  the  100-watt  transmitter  has  to  be  cut 
down  to  the  15  watt  driving  level.  This  is  a 
nuisance,  as  it  requires  the  addition  of  swamp- 
ing resistors  to  the  output  circuit  of  the  trans- 
mitter-exciter. The  triode  tubes,  therefore, 
would  lend  themselves  to  a much  more  con- 
venient driving  arrangement  than  would  the 
tetrode  tubes,  simply  because  their  grid  drive 
requirements  fall  within  the  power  output 
range  of  the  exciter  unit. 

On  the  other  hand,  if  the  transmitter-exciter 
output  level  is  of  the  order  of  15-40  watts 
(the  Johnson  Ranger,  for  example)  sufficient 
drive  for  triode  tubes  tunning  1 -kilowatt  input 
would  be  lacking.  Tetrode  tubes  requiring  low 
grid  driving  power  would  have  to  be  employed 
in  a high-level  stage,  or  smaller  triode  tubes  re- 
quiring modest  grid  drive  and  running  250 
watts  or  so  would  have  to  be  used. 

Power  Amplifier  Either  push-pull  or  single 
Design — Choice  ended  circuits  may  be  em- 

of  Circuits  ployed  in  the  power  ampli- 

fier. Using  modern  tubes 
and  properly  designed  circuits,  either  type  is 
capable  of  high  efficiency  operation  and  low 
harmonic  output.  Push-pull  circuits,  whether 
using  triode  or  tetrode  tubes  usually  employ 
link  coupling  between  the  amplifier  stage  and 
the  feed  line  running  to  the  antenna  or  the  an- 
tenna tuner. 

It  is  possible  to  use  the  link  circuit  in  either 
an  unbalanced  or  balanced  configuration,  as 
shown  in  figure  1,  using  unbalanced  coaxial 
line,  or  balanced  twin-line. 


Figure  2 

CONVENTIONAL  PUSH-PULL 
AMPLIFIER  CIRCUIT 

The  mechanical  layout  should  be  symmetrical 

and  the  output  coupling  provision  must  be 

evenly  balanced  with  respect  to  the  plate  coil 

Cj — Approx.  1.S  ixjifd.  per  meter  of  wave- 
length per  section 

Ci — Refer  to  plate  tank  capacitor  design  in 
Chapter  1 1 

Ci — May  be  500  iiixfd.,  I0,000-vo/t  type  cer- 
amic capacitor 

NC — Max.  usable  capacitance  should  be  great- 
er, and  min.  capacitance  less  than  rated 
grid-plate  capacity  of  tubes  in  amplifier. 
50%  greater  air  gap  than  Ct. 

R]—100  ohms,  20  wotfs.  This  resistor  serves 
as  low  Q r-f  choke. 

RFCj — All-band  r-f  choke  suitable  for  plate 
current  of  tubes 

Mj-Mr^Suitable  meters  for  d-c  grid  and  plate 
currents 

All  low  voltage  .001  nfd.  and  .01  jifd.  by-pass 
capacitors  are  ceramic  disc  units  (Centra- 
lab  DD  or  equiv.) 

Li — SO-watt  plug-in  coil,  center  link 

Li — Plug-in  coil,  center  link,  of  suitable  power 
rating, 

G>nimon  technique  is  to  employ  plug-in 
plate  coils  with  the  push-pull  amplifier  stage. 
This  necessitates  some  kind  of  opening  for  coil 
changing  purposes  in  the  "electrically  tight” 
enclosure  surrounding  the  amplifier  stage.  Care 
must  be  used  in  the  design  and  construction 
of  the  door  for  this  opening  or  leakage  of  har- 
monics through  the  opening  will  result,  with 
the  attendant  TVI  problems. 

Single  ended  amplifiers  may  also  employ 
link-coupled  output  devices,  although  the  trend 
is  to  use  pi-network  circuits  in  conjunction 
with  single  ended  tetrode  stages.  A tapped  or 
otherwise  variable  tank  coil  may  be  used  which 
is  adjustable  from  the  front  panel,  eliminating 
the  necessity  of  plug-in  coils  and  openings 
into  the  shielded  enclosure  of  the  amplifier. 
Pi-network  circuits  are  becoming  increasingly 
popular  as  coaxial  feed  systems  are  coming 
into  use  to  couple  the  output  circuits  of  trans- 
mitters directly  to  the  antenna. 


612  H.F.  Power  Amplifiers 


THE  RADIO 


29-2  Push-Pull  Triode 

Amplifiers 

Figure  2 shows  a basic  push-pull  triode 
amplifier  circuit.  While  variations  in  the  meth- 
od of  applying  plate  and  filament  voltages  and 
bias  are  sometimes  found,  the  basic  circuit 
remains  the  same  in  all  amplifiers. 

Filament  Supply  The  amplifier  filament  trans- 
former should  be  placed 
right  on  the  amplifier  chassis  in  close  proximity 
to  the  tubes.  Short  filament  leads  are  necessary 
to  prevent  excessive  voltage  drop  in  the  con- 
necting leads,  and  also  to  prevent  r-f  pickup 
in  the  filament  circuit.  Long  filament  leads 
can  often  induce  instability  in  an  otherwise 
stable  amplifier  circuit,  especially  if  the  leads 
are  exposed  to  the  radiated  field  of  the  plate 
circuit  of  the  amplifier  stage.  The  filament 
voltage  should  be  the  correct  value  specified 
by  the  tube  manufacturer  when  measured  at  the 
tube  sockets.  A filament  transformer  having  a 
tapped  primary  often  will  be  found  useful  in 
adjusting  the  filament  voltage.  When  there  is 
a choice  of  having  the  filament  voltage  slight- 
ly higher  or  slightly  lower  than  normal,  the 
higher  voltage  is  preferable.  If  the  amplifier  is 
to  be  overloaded,  a filament  voltage  slight- 
ly higher  than  the  rated  value  will  give  greater 
tube  life. 

Filament  bypass  capacitors  should  be  low  in- 
ternal inductance  units  of  approximately  .01 
gfd.  A separate  capacitor  should  be  used  for 
each  socket  terminal.  Lower  values  of  capaci- 
tance should  be  avoided  to  prevent  spurious 
resonances  in  the  internal  filament  structure  of 
the  tube.  Use  heavy,  shielded  filament  leads  for 
low  voltage  drop  and  maximum  circuit  isola- 
tion. 

Plate  Feed  The  series  plate  voltage  feed 
shown  in  figure  2 is  the  most 
satisfactory  method  for  push-pull  stages.  This 
method  of  feed  puts  high  voltage  on  the  plate 
tank  coil,  but  since  the  r-f  voltage  on  the  coil 
is  in  itself  sufficient  reason  for  protecting  the 
coil  from  accidental  bodily  contact,  no  addi- 
tional protective  arrangements  are  made  neces- 
sary by  the  use  of  series  feed. 

The  insulation  in  the  plate  supply  circuit 
should  be  adequate  for  the  voltages  encoun- 
tered. In  general,  the  insulation  should  be 
rated  to  withstand  at  least  four  times  the  max- 
imum d-c  plate  voltage.  For  safety,  the  plate 
meter  should  be  placed  in  the  cathode  return 
lead,  since  there  is  danger  of  voltage  break- 
down between  a metal  panel  and  the  meter 


movement  at  plate  voltages  much  higher  than 

one  thousand. 

Grid  Bias  The  recommended  method  of  ob- 
taining bias  for  c-w  or  plate  mod- 
ulated telephony  is  to  use  just  sufficient  fixed 
bias  to  protect  the  tubes  in  the  event  of  ex- 
citation failure,  and  to  obtain  the  rest  by  the 
voltage  drop  caused  by  flow  of  rectified  grid 
current  through  a grid  resistor.  If  desired,  the 
bias  supply  may  be  omitted  for  telephony  if  an 
overload  relay  is  incorporated  in  the  plate  cir- 
cuit of  the  amplifier,  the  relay  being  adjusted 
to  trip  immediately  when  excitation  is  re- 
moved from  the  stage. 

The  grid  resistor  Ri  serves  effectively  as 
an  r-f  choke  in  the  grid  circuit  because  the  im- 
pressed r-f  voltage  is  low,  and  the  Q of  the 
resistor  is  poor.  No  r-f  choke  need  be  used  in 
the  grid  bias  return  lead  of  the  amplifier,  other 
than  those  necessary  for  harmonic  suppression. 

The  bias  supply  may  be  built  upon  the  am- 
plifier chassis  if  care  is  taken  to  prevent  r-f 
from  finding  its  way  into  the  supply.  Ample 
shielding  and  lead  filtering  must  be  employed 
for  sufficient  isolation. 

The  Grid  Circuit  As  the  power  in  the  grid 
circuit  is  much  lower  than 
in  the  plate  circuit,  it  is  customary  to  use  a 
close-spaced  split-stator  grid  capacitor  with 
sufficient  capacitance  for  operation  on  the 
lowest  frequency  band.  A physically  small  ca- 
pacitor has  a greater  ratio  of  maximum  to 
minimum  capacitance,  and  it  is  possible  to  ob- 
tain a unit  that  will  be  sarisfaaory  on  all  bands 
from  10  to  80  meters  without  the  need  for  aux- 
iliary padding  capacitors.  The  rotor  of  the  grid 
capacitor  is  grounded,  simplifying  mounting  of 
the  capacitor  and  providing  circuit  balance  and 
electrical  symmetry.  Grounding  the  rotor  also 
helps  to  retard  v-h-f  parasitics  by  by-passing 
them  to  ground  in  the  grid  circuit.  The  L/C 
ratio  in  the  grid  circuit  should  be  fairly  low, 
and  care  should  be  taken  that  circuit  reso- 
nance is  not  reached  with  the  grid  capacitor  at 
minimum  capacitance.  That  is  a direct  invita- 
tion for  instability  and  parasitic  oscillations 
in  the  stage.  The  grid  coil  may  be  wound  of 
no.  14  wire  for  driving  powers  of  up  to  100 
watts.  To  restrict  the  field  and  thus  aid  in 
neutralizing,  the  grid  coil  should  be  physical- 
ly no  larger  than  absolutely  necessary. 

Circuit  Layout  The  most  important  consid- 
eration in  constructing  a 
push-pull  amplifier  is  to  maintain  elearical 
symmetry  on  both  sides  of  the  balanced  clr- 


HANDBOOK 


Design  613 


cuit.  Of  utmost  importance  in  maintaining  elec- 
trical balance  is  the  control  of  stray  capaci- 
tance between  each  side  of  the  circuit  and 
ground. 

Large  masses  of  metal  placed  near  one  side 
of  the  grid  or  plate  circuits  can  cause  serious 
unbalance,  especially  at  the  higher  frequen- 
cies, where  the  tank  capacitance  between  one 
side  of  the  tuned  circuit  and  ground  is  often 
quite  small  in  itself.  Capacitive  unbalance 
most  often  occurs  when  a plate  or  grid  coil  is 
located  with  one  of  its  ends  close  to  a metal 
panel.  The  solution  to  this  difficulty  is  to 
mount  the  coil  parallel  to  the  panel  to  make 
the  capacitance  to  ground  equal  from  each  end 
of  the  coil,  or  to  place  a grounded  piece  of 
metal  opposite  the  "free”  end  of  the  coil  to 
accomplish  a capacity  balance. 

Whenever  possible,  the  grid  and  plate  coils 
should  be  mounted  at  right  angles  to  each 
other,  and  should  be  separated  fat  enough 
apart  to  reduce  coupling  between  them  to  a 
minimum.  Coupling  between  the  grid  and  plate 
coils  will  tend  to  make  neutralization  frequency 
sensitive,  and  it  will  be  necessary  to  readjust 
the  neutralizing  capacitors  of  the  stage  when 
changing  bands. 

All  t-f  leads  should  be  made  as  short  and 
direct  as  possible.  The  leads  from  the  tube 
grids  or  plates  should  be  conneaed  directly 
to  their  respective  tank  capacitors,  and  the 
leads  between  the  tank  capacitors  and  coils 
should  be  as  heavy  as  the  wire  that  is  used 
in  the  coils  themselves.  Plate  and  grid  leads 
to  the  tubes  may  be  made  of  flexible  tinned 
braid  or  flat  copper  strip.  Neutralizing  leads 
should  run  directly  to  the  tube  grids  and  plates 
and  should  be  separate  from  the  grid  and  plate 
leads  to  the  tank  circuits.  Having  a portion  of 
the  plate  or  grid  connections  to  their  tank  cir- 
cuits serve  as  part  of  a neutralizing  lead  can 
often  result  in  amplifier  instability  at  certain 
operating  frequencies. 

Excitation  In  general  it  may  be  stated 

Requirements  that  the  overall  power  require- 
ment for  grid  circuit  excitation 
to  a push-pull  triode  amplifier  is  approximately 
10  per  cent  of  the  amount  of  the  power  output 
of  the  stage.  Tetrodes  require  about  1 per  cent 
to  3 percent  excitation,  referred  to  the  power 
output  of  the  stage.  Excessive  excitation  to 
pentodes  or  tetrodes  will  often  result  in  re- 
duced power  output  and  efficiency. 

Push-Pull  Symmetry  is  the  secret  of  suc- 

Amplifier  cessful  amplifier  design.  Shown 
Construction  in  figure  3 is  rhe  top  view  of 
a 350  watt  push-pull  all  band 


Figure  3 

LAYOUT  OF  350-WATT  PUSH-PULL 
TRIODE  AMPLIFIER 
Two  81] -A  tubes  are  employed  in  this  circuit. 
Plate  tuning  capacitor  is  at  left  of  chassis, 
with  swinging-link  type  plug-in  coil  assembly 
mounted  above  it  Rotor  of  split-stator  capaci- 
tor may  be  insulated  from  ground  to  increase 
voltage  breakdown  rating  of  capacitor.  Note 
that  pickup  link  is  series-tuned  to  reduce  cir- 
cuit reactance.  One  corner  of  rotor  plate  of 
series  capacitor  is  bent  so  that  capacitor 
shorts  itself  out  at  maximum  capacitance. 
Grid  circuit  coil  and  capacitor  are  at  right. 
Center-linked  plug-in  coil  is  employed.  Para- 
sitic chokes  are  placed  in  grid  leads  adjacent 
to  the  tube  sockets,  and  tube  filaments  are 
bypassed  to  ground  with  .01  fifd.  ceramic 
capacitors.  Complete  area  above  the  chassis 
is  enclosed  with  perforated  screen  to  reduce 
radiation  of  r.f.  energy. 


amplifier  employing  811-A  tubes.  The  circuit 
corresponds  to  that  shown  in  figure  2 except 
that  the  811 -As  are  zero  bias  tubes.  The  bias 
terminals  of  the  circuit  are  therefore  jumpered 
together  and  no  external  bias  supply  is  re- 
quired at  plate  potentials  less  than  1300  volts. 

All  r-f  components  are  mounted  above 
deck.  The  plate  circuit  tuning  capacitor  and 
swinging  link  tank  coil  are  to  the  left,  with 
the  two  disc-type  neutralizing  capacitors  be- 
tween the  tank  circuit  and  the  tubes.  At  the 
right  of  the  chassis  is  the  grid  tank  circuit. 
Small  parasitic  chokes  may  be  seen  between 
the  tube  sockets  and  the  grid  circuit.  Plate 
and  grid  meters  are  placed  in  the  under-chassis 
area  where  they  are  shielded  from  the  r-f  field 
of  the  amplifier. 

Larger  triode  mbes  such  as  the  810  and 
8000  make  excellent  r-f  amplifiers  at  the  kilo- 
watt level,  but  care  must  be  taken  in  ampli- 
fier layout  as  the  inter-electrode  capacitance  of 
these  tubes  is  quite  high.  One  tube  and  one 
neutralizing  capacitor  is  placed  on  each  side 
of  the  tank  circuit  (figures  4 and  5)  to  permit 
very  short  interconnecting  leads.  The  relative 
position  of  the  tubes  and  capacitors  is  trans- 


614  H.F.  Power  Amplifiers 


THE  RADIO 


Figure  4 

UNIQUE  CHASSIS  LAYOUT  PERMITS 
SHORT  LEADS  IN  KILOWATT 
AMPLIFIER 

Lprge  size  components  required  for  high  level 
amplifier  often  complicate  amplifier  layout. 
In  this  design,  the  plate  tank  capacitor  sits 
astride  small  chassis  running  lengthwise  on 
main  chassis.  Inductor  is  mounted  to  phenolic 
plate  atop  capacitor.  Variable  link  is  panel 
driven  through  right-angle  gear  drive.  Plate 
circuit  is  grounded  by  safety  arm  when  panel 
door  is  opened.  Note  that  plate  capacitor  is 
mounted  on  four  TV-type  capacitors  which 
serve  to  bypass  unit,  and  also  act  as  supports. 
A small  parasitic  choke  is  visible  next  to  the 
grid  terminal  of  the  810  tube. 


posed  on  each  side  of  the  chassis,  as  shown  in 
the  illustrations.  The  plate  tank  coil  is  mount- 
ed parallel  to  the  front  panel  of  the  amplifier 
on  a phenolic  plate  supported  by  the  tuning 
capacitor  which  sits  atop  a small  chassis-type 
box.  The  grid  circuit  tuning  capacitor  is  located 
within  this  box,  as  seen  in  figure  6.  An  ex- 
ternal bias  supply  is  required  for  proper  ampli- 
fier operation.  Operating  voltages  may  be  de- 
termined from  the  instruction  sheets  for  the 
particular  tube  to  be  employed. 

Whenever  the  amplifier  enclosure  requires 
a panel  door  for  coil  changing  access  it  is  wise 
to  place  a power  interlock  on  the  door  that 
will  turn  off  the  high  voltage  supply  whenever 
the  door  is  open! 


Figure  5 

LEFT-HAND  VIEW  OF  KILOWATT 
AMPLIFIER  OF  FIGURE  4 

Above  shielded  meter  box  is  the  protective 
'*micre~switch''  which  opens  the  primary  power 
circuit  when  the  panei  door  is  not  dosed.  Tube 
sockets  are  recessed  in  the  chassis  so  that 
top  of  tube  socket  sheiis  are  about  i/2-incb 
above  chassis  ievei.  On  right  side  of  ampiifier 
(facing  it  from  the  rear)  the  tube  socket  is 
nearest  the  panei,  with  the  neutraiizing  co- 
pacitor  behind  it.  On  the  opposite  side,  the 
capacitor  is  nearest  the  panei  with  the  tube 
directly  behind  it.  This  layout  transposition 
produces  very  short  neutralizing  leads,  since 
connections  may  be  made  through  the  stator 
of  plate  tuning  capacitor. 

29-3  Push-Pull 

Tetrode  Amplifiers 

Tetrode  tubes  may  be  employed  in  push-pull 
amplifiers,  although  the  modern  trend  is  to 
parallel  operation  of  these  tubes.  A typical 
circuit  for  push-pull  operation  is  shown  in 
figure  7.  The  remarks  concerning  the  filament 
supply,  plate  feed,  and  grid  bias  in  Section 
29-2  apply  equally  to  tetrode  stages.  Because 
of  the  high  circuit  gain  of  the  tetrode  ampli- 
fier, extreme  care  must  be  taken  to  limit 
interstage  feedback  to  an  absolute  minimum. 
Many  amateurs  have  had  bad  luck  with  tet- 
rode tubes  and  have  been  plagued  with  para- 
sitics  and  spurious  oscillations.  It  must  be  re- 
membered with  high  gain  tubes  of  this  type 


HANDBOOK 


P-P  Tetrode  Amplifier  615 


Figure  6 

UNDER  CHASSIS  VIEW  OF 
1 -KILOWATT  TRIODE  AMPLIFIER 

The  grid  circuit  tuning  capacitor  and  plate 
circuit  r-f  choke  are  contained  in  the  below 
chassis  enclosure  formed  by  a small  chassis 
mounted  at  right  angles  to  the  front  panel. 
The  bandswitch  coil  assembly  for  the  grid 
circuit  is  mounted  on  two  brackets  above  this 
cutout.  A metal  screen  attached  to  the  bottom 
of  the  amplifier  completes  the  TVI-proof 
enclosure. 


that  almost  full  output  can  be  obtained  with 
practically  zero  grid  excitation.  Any  minute 
amount  of  energy  fed  back  from  the  plate  cir- 
cuit to  the  grid  circuit  can  cause  instability  or 
oscillation.  Unless  suitable  precautions  are  in- 
corporated in  the  electrical  and  mechanical  de- 
sign of  the  amplifier,  this  energy  feedback  will 
inevitably  occur. 

Fortunately  these  precautions  are  simple. 
The  grid  and  filament  circuits  must  be  isolated 
from  the  plate  circuit.  This  is  done  by  placing 
these  circuits  in  an  "electrically  tight”  box. 
All  leads  departing  from  this  box  are  by-pass- 
ed and  filtered  so  that  no  r^  energy  can  pass 
along  the  leads  into  the  box.  This  restricts  the 
energy  leakage  path  between  the  plate  and  grid 
circuits  to  the  residual  plate-to-grid  capacity 
of  the  tetrode  tubes.  This  capacity  is  of  the 
order  of  0.25  g/xfd.  per  tube,  and  under  normal 
conditions  is  sufficient  to  produce  a highly  re- 
generative condition  in  the  amplifier.  Whether 
or  not  the  amplifier  will  actually  break  into  os- 
cillation is  dependent  upon  circuit  losses  and 
residual  lead  inductance  of  the  stage.  Suffice 
to  say  that  unless  the  tubes  are  actually  neu- 
tralized a condition  exists  that  will  lead  to 
circuit  instability  and  oscillation  under  cer- 
tain operating  conditions.  With  luck,  and  a 


Figure  7 

CONVENTIONAL  PUSH-PULL 
TETRODE  AMPLIFIER  CIRCUIT 

Push-pull  amplifier  uses  many  of  the  same 
components  required  by  triode  tubes  (see 
figure  2).  Screen  supply  is  also  required. 
B — Blower  for  filament  seals  of  tubes. 

C.< — Low  internal  inductance  capacitor,  .001 
tifd.,  SKY.  Centralab  type  8S8S  - 1000. 
NC— See  text  and  figure  8. 

PC^Parasitic  choke.  SO  ohm,  2-wott  compo- 
sition resistor  wound  with  3 turns  till  e. 
wire. 

Note:  Strap  multiple  screen  terminals  together 
at  socket  with  copper  ribbon.  Attach 
PC  to  center  of  strap. 


heavily  loaded  plate  circuit,  one  might  be  able 
to  use  an  un-neutralized  push-pull  tetrode  am- 
plifier stage  and  suffer  no  ill  effects  from  the 
residual  grid-plate  feedback  of  the  tubes.  In 
fact,  a minute  amount  of  external  feedback  in 
the  power  leads  to  the  amplifier  may  just  (by 
chance)  cancel  out  the  inherent  feedback  of 
the  amplifier  circuit.  Such  a condition,  how- 
ever, results  in  an  amplifier  that  is  not  "re- 
produceable.”  There  is  no  guarantee  that  a du- 
plicate amplifier  will  perform  in  the  same,  sta- 
ble manner.  This  is  the  one,  great  reason  that 
many  amateurs  having  built  a tetrode  amplifier 
that  "looks  just  like  the  one  in  the  book”  find 
out  to  their  sorrow  that  it  does  not  "work  like 
the  one  in  the  book.” 

This  borderline  situation  can  easily  be  over- 
come by  the  simple  process  of  neutralizing  the 
high-gain  tetrode  tubes.  Once  this  is  done,  and 
the  amplifier  is  tested  for  parasitic  oscilla- 
tions (and  the  oscillations  eliminated  if  they 
occur)  the  tetrode  amplifier  will  perform  in  an 
excellent  manner  on  all  bands.  In  a word,  it 
will  be  "reproduceable.” 


616  H.F.  Power  Amplifiers 


THE  RADIO 


Figure  8 

REAR  VIEW  OF  PUSH  PULL 
4-250A  AMPLIFIER 

The  neutralizing  rods  are  mounted  on  ceramic 
feedthrough  insulators  adjacent  to  each  tube 
socket.  Low  voltage  power  leads  leave  the 
grid  circuit  compartment  via  Hypass  capaci- 
tors located  on  the  lower  left  corner  of  the 
chassis.  A screen  plate  covers  the  rear  of  the 
amplifier  during  operation.  This  plate  was 
removed  for  the  photograph. 


As  a summation,  three  requirements  must 
be  met  for  proper  operation  of  tetrode  tubes' — 
whether  in  a push-pull  or  parallel  mode; 

1.  Complete  isolation  must  be  achieved  be- 
tween the  grid  and  plate  circuits. 

2.  The  tubes  must  be  neutralized. 

3.  The  circuit  must  be  parasitic-free. 

Amplifier  The  push-pull  tetrode  ampli- 

Construction  fier  should  be  built  around 

two  "r-f  tight”  boxes  for  the 
grid  and  plate  circuits.  A typical  layout  that  has 
proven  very  satisfactory  is  shown  in  figures  8 
and  9.  The  amplifier  is  designed  around  a Bar- 
ker & Williamson  "butterfly"  tuning  capacitor. 
The  4-250A  tetrode  tubes  are  mounted  at  the 
rear  of  the  chassis  on  each  side  of  the  capaci- 
tor. The  base  shells  of  the  tubes  are  grounded 
by  spring  clips,  and  short  adjustable  rods  pro- 
ject up  beside  each  tube  to  act  as  neutralizing 
capacitors.  The  leads  to  these  rods  are  cross- 
connected  beneath  the  chassis  and  the  rods 
provide  a small  value  of  capacitance  to  the 
plates  of  the  tubes.  This  neutralization  is  nec- 
essary when  the  tube  is  operated  with  high 


Figure  9 

UNDER  CHASSIS  VIEW  OF 
4-250A  AMPLIFIER 

The  bias  supply  for  the  amplifier  is  mounted 
at  the  front  of  the  chassis  between  the  two 
control  shafts.  A blower  motor  is  mounted 
beneath  each  tube  socket.  A screened  plate 
is  placed  on  the  bottom  of  the  chassis  to  com- 
plete the  under-chassis  shielding. 


power  gain  and  high  screen  voltage.  As  the 
operating  frequency  of  the  tube  is  increased, 
the  inductance  of  the  internal  screen  support 
lead  of  the  tube  becomes  an  important  part  of 
the  screen  ground  return  circuit.  At  some  criti- 
cal frequency  (about  45  Me.  for  the  4-250A 
tube)  the  screen  lead  inductance  causes  a 
series  resonant  condition  and  the  tube  is  said 
to  be  "self-neutralized”  at  this  frequency. 
Above  this  frequency  the  screen  of  the  tetrode 
tube  cannot  be  held  at  ground  potential  by  the 
usual  screen  by-pass  capacitors.  With  normal 
circuitry,  the  tetrode  tube  will  have  a tendency 
to  self-oscillate  somewhere  in  the  120  Me.  to 
160  Me.  region.  Low  capacity  tetrodes  that  can 
operate  efficiently  at  such  a high  frequency  are 
capable  of  generating  robust  parasitic  oscilla- 
tions in  this  region  while  the  operator  is  vain- 
ly trying  to  get  them  operating  at  some  lower 
frequency.  The  solution  is  to  introduce  enough 
loss  in  the  circuit  at  the  frequency  of  the 
parasitic  so  as  to  render  oscillation  impos- 
sible. This  procedure  has  been  followed  in  this 
amplifier. 

During  a long  series  of  experiments  de- 
signed to  stabilize  large  tetrode  tubes,  it  was 
found  that  suppression  circuits  were  most  ef- 
fective when  inserted  in  the  screen  lead  of  the 


HANDBOOK 


Pi-Network  Amplifiers  617 


tetrode.  The  screen,  it  seemed,  would  have  r-f 
potentials  measuring  into  the  thousands  of  volts 
upon  it  during  a period  of  parasitic  oscillation. 
By-passing  the  screen  to  ground  with  copper 
strap  connections  and  multiple  by-pass  capaci- 
tors did  little  to  decrease  the  amplitude  of  the 
oscillation.  Excellent  parasitic  suppression  was 
brought  about  by  strapping  the  screen  leads 
of  the  4-250A  socket  together  (figure  7)  and 
inserting  a parasitic  choke  between  the  screen 
terminal  of  the  socket  and  the  screen  by-pass 
capacitor. 

After  this  was  done,  a very  minor  tendency 
towards  self-oscillation  was  noted  at  extreme- 
ly high  plate  voltages.  A small  parasitic  choke 
in  each  grid  lead  of  the  4-250A  tubes  elimi- 
nated this  completely. 

The  neutralizing  rods  are  mounted  upon  two 
feedthrough  insulators  and  cross-connected  to 
the  4-250A  control  grids  beneath  the  chassis. 
These  tods  are  threaded  so  that  they  may  be 
run  up  and  down  the  insulator  bolt  for  neu- 
tralizing adjustment. 

Because  of  the  compact  size  of  many  tetrodes 
it  is  necessary  to  cool  the  filament  seals  of  the 
tube  with  a blast  of  ait.  A small  blower  can  be 
mounted  beneath  the  chassis  to  project  cooling 
air  directly  at  the  socket  of  the  tube  as  shown 
in  figure  9. 

Inductive  Tuning  of  The  plate  tank  circuit 
Push-Pull  Amplifiers  of  the  push-pull  ampli- 
fier must  have  a low 
impedance  to  ground  at  harmonic  frequencies 
to  provide  adequate  harmonic  suppression.  The 
usual  split-stator  tank  capacitor,  however,  has 
an  uncommonly  high  impedance  in  the  VHF 
region  wherein  the  interference-causing  har- 
monics lie.  A push-pull  vacuum-type  capacitor 
may  be  used  as  these  units  have  very  low  in- 
ternal inductance,  but  the  cost  of  such  a capac- 
itor is  quite  high. 

A novel  solution  to  this  problem  is  to  em- 
ploy a split  stator  capacitor  made  up  of  two 
inexpensive  fixed  vacuum  capacitors.  Ampli- 
fier adjustment  can  then  best  be  accomplished 
by  inductive  tuning  of  the  plate  tank  coil  as 


Figure  10 

INDUCTIVE  TUNING  MAY  BE 
EMPLOYED  IN  HIGH  POWER 
AMPLIFIER 

Two  fixed  vacuum  capacitors  form  split-stator 
capacitance,  providing  very  low  inductance 
ground  path  for  plate  circuit  harmonics,  Tun~ 
ing  is  accomplished  by  means  of  shorted, 
single-turn  link  placed  in  center  of  tank 
coil.  Shorted  link  is  made  from  %-inch  section 
cut  from  copper  water  pipe.  Larger  link  out- 
side of  tank  coil  is  antenna  pick-up  coil. 


seen  in  figure  10.  Two  fixed  vacuum  capaci- 
tors are  mounted  vertically  upon  the  chassis 
and  the  upper  terminals  are  attached  to  the 
plates  of  the  amplifier  tubes  by  means  of  low 
impedance  straps.  Resonance  is  established  by 
rotation  of  a shorted  copper  loop  located  with- 
in the  amplifier  tank  coil.  This  loop  is  made 
of  a %"  long  section  of  copper  water  pipe, 
two  inches  in  diameter.  Approximate  resonance 
is  established  by  varying  the  spacing  between 
the  turns  of  the  copper  tubing  tank  coil.  In- 
ductive coupling  is  used  between  the  tank  coil 
and  the  antenna  circuit  in  the  usual  manner. 
Sufficient  range  to  enable  the  operator  to  cover 
a complete  high  frequency  band  may  be  had 
with  this  interesting  tuning  method. 

29-4  Tetrode  Pi- 

Network  Amplifiers 

The  most  popular  amplifier  today  for  both 
commercial  and  amateur  use  is  the  pi-network 
configuration  shown  in  figure  11.  This  circuit 
is  especially  suited  to  tetrode  mbes,  although 
triode  tubes  may  be  used  under  certain  circum- 
stances. 

A common  form  of  pi-network  amplifier  is 
shown  in  figure  11  A.  The  pi  circuit  forms  the 
matching  system  between  the  plate  of  the  am- 
plifier tube  and  the  low  impedance,  unbalanced 
antenna  circuit.  The  coil  and  input  capacitor 


618  H.F.  Power  Amplifiers 


THE  RADIO 


— BIAS  -f-  1 ISV.'V 


Figure  1 1 

TYPICAL  PI-NETWORK  CONFIGURATIONS 

A — Split  grid  circuit  provides  out-of-phase  voltage  for  grid  neutralization  of  tetrode  tube.  Rotary  coil 
is  employed  in  plate  circuit,  with  small,  fixed  auxiliary  coil  for  28  Me.  Multiple  tuning  grid  tank  Ti 
covers  3.5  - 30  Me.  without  switching. 

B — Tapped  grid  and  plate  inductors  are  used  with  "bridge  type"  neutralizing  circuit  for  tetrode  ampli- 
fier stage.  Vacuum  tuning  capacitor  is  used  in  input  section  of  pi-network. 

C — Untuned  input  circuit  (resistance  loaded)  and  plate  inductor  ganged  with  tuning  capacitor  comprise 
simple  amplifier  configuration. 

PCj,  PCi — 57  ohm,  2 watt  composition  resistor,  wound  with  3 turns  #18  c.  wire. 


HANDBOOK 


Pi-Network  Amplifiers  619 


of  the  pi  may  be  varied  to  tune  the  circuit  over 
a 10  to  1 frequency  range  (usually  3.0-30 
Me. ) . Operation  over  the  20  - 30  Me.  range 
takes  place  when  the  variable  slider  on  coil  Li 
is  adjusted  to  short  this  coil  out  of  the  circuit. 
Coil  Li  therefore  comprises  the  tank  inductance 
for  the  highest  portion  of  the  operating  range. 
This  coil  has  no  taps  or  sliders  and  is  con- 
structed for  the  highest  possible  Q at  the  high 
frequency  end  of  the  range.  The  adjustable 
coil  (because  of  the  variable  tap  and  physical 
construction)  usually  has  a lower  Q than  that 
of  the  fixed  coil. 

The  degree  of  loading  is  controlled  by  ca- 
pacitors Cj  and  Cr,.  The  amount  of  circuit  ca- 
pacity requited  at  this  point  is  inversely  pro- 
portional to  the  operating  frequency  and  to 
the  impedance  of  the  antenna  circuit.  A load- 
ing capacitor  range  of  100  ggfd.  to  2500 
g/ifd.  is  normally  ample  to  cover  the  3.5  - 30 
Me.  range. 

The  pi  circuit  is  usually  shunt-fed  to  remove 
the  d.c.  plate  voltage  from  the  coils  and  capa- 
citors. The  components  are  held  at  ground  po- 
tential by  completing  the  circuit  ground 
through  the  choke  RFCi.  Great  stress  is  placed 
upon  the  plate  circuit  choke  RFG.  This  com- 
ponent must  be  specially  designed  for  this 
mode  of  operation,  having  low  inter-turn  ca- 
pacity and  no  spurious  internal  resonances 
throughout  the  operating  range  of  the  ampli- 
fier. 

Parasitic  suppression  is  accomplished  by 
means  of  chokes  PC-1  and  PC-2  in  the  screen 
and  grid  leads  of  the  tetrode.  Suitable  values 
for  these  chokes  are  given  in  the  parts  list  of 
figure  11.  Effective  parasitic  suppression  is  de- 
pendent to  a large  degree  upon  the  choice  of 
screen  bypass  capacitor  Ci.  This  component 
must  have  extremely  low  inductance  through- 
out the  operating  range  of  the  amplifier  and 
well  up  into  the  VHF  parasitic  range.  The  ca- 
pacitor must  have  a voltage  rating  equal  to  at 
least  twice  the  screen  potential  ( four  times  the 
screen  potential  for  plate  modulation).  There 
are  practically  no  capacitors  available  that  will 
perform  this  difficult  task.  One  satisfactory  so- 
lution is  to  allow  the  amplifier  chassis  to  form 
one  plate  of  the  screen  capacitor.  A "sandwich” 
is  built  upon  the  chassis  with  a sheet  of  insu- 
lating material  of  high  dielectric  constant  and 
a matching  metal  sheet  which  forms  the  screen 
side  of  the  capacitance.  A capacitor  of  this 
type  has  very  low  internal  inductance  but  is 
very  bulky  and  takes  up  valuable  space  be- 
neath the  chassis.  One  suitable  capacitor  for 
this  position  is  the  Centralab  type  858S-1000, 


EXCITER 4r AMPLIFIER 


CAPACITIVE  COUPLING  CIRCUIT 
BETWEEN  EXCITER  AND  FINAL 
AMPLIFIER  STAGE  MAKES  USE 
OF  ONE  COMMON  TANK  CIRCUIT 

Grid  circuit  of  ampiifier  stage  serves  as  piate 
circuit  of  exciter  in  this  simplified  circuit. 
Leads  between  the  tubes  must  be  kept  short 
and  direct. 


rated  at  1000  ggfd.  at  5000  volts.  This  com- 
pact ceramic  capacitor  has  relatively  low  inter- 
nal inductance  and  may  be  mounted  to  the 
chassis  by  a 6-32  bolt.  It  is  shown  in  various 
amplifiers  described  in  this  chapter.  Further 
screen  isolation  may  be  provided  by  a shielded 
power  lead,  isolated  from  the  screen  by  a .001 
gfd.  ceramic  capacitor  and  a 100  ohm  car- 
bon resistor. 

Various  forms  of  the  basic  pi-network  am- 
plifier are  shown  in  figure  11.  The  A config- 
uration employs  the  so-called  "all-band”  grid 
tank  circuit  and  a rotary  pi-network  coil  in  the 
plate  circuit.  The  B circuit  uses  coil  switching 
in  the  grid  circuit,  bridge  neutralization,  and 
a tapped  pi-network  coil  with  a vacuum  tun- 
ing capacitor.  Figure  IIC  shows  an  interesting 
circuit  that  is  becoming  more  popular  for 
class  ABl  linear  operation.  A tetrode  tube  op- 
erating under  class  ABl  conditions  draws  no 
grid  current  and  requires  no  grid  driving 
power.  Only  r-f  voltage  is  required  for  proper 
operation.  It  is  possible  therefore  to  dispense 
with  the  usual  tuned  grid  circuit  and  neutraliz- 
ing capacitor  and  in  their  place  employ  a sim- 
ple load  resistor  in  the  grid  circuit  across 
which  the  required  excitation  voltage  may  be 
developed.  This  resistor  can  be  of  the  order  of 
50  - 300  ohms,  depending  upon  circuit  re- 
quirements. Considerable  power  must  be  dis- 
sipated in  the  resistor  to  develop  sufficient 
grid  swing,  but  driving  power  is  often  cheaper 
to  obtain  than  the  cost  of  the  usual  grid  cir- 
cuit components.  In  addition,  the  low  im- 
pedance grid  return  removes  the  tendency  to- 
wards instability  that  is  so  common  to  the 
circuits  of  figure  llA  and  IIB.  Neutralization 
is  not  required  of  the  circuit  of  figure  IIC, 
and  in  many  cases  parasitic  suppression  may 
be  omitted.  The  price  that  must  be  paid  is  the 


620  H.F.  Power  Amplifiers 


THE  RADIO 


Figure  13 

MIDGET  LINEAR  AMPLIFIER  FOR 
MOBILE  SIDEBAND  OPERATION 
FEATURES  WATER  COOLED 
4W-300B  TETRODE  TUBES 

Miniature  amplifier  is  comparable  in  size  with 
sideband  exciter.  Antenna,  plate,  and  screen 
current  are  monitored  by  panel  lamps  used  in 
place  of  larger  and  more  expensive  meters. 
Inductor  slug  of  grid  coil  is  located  below 
plate  circuit  tuning  capacitor. 


additional  excitation  that  is  required  to  de- 
velop operating  voltage  across  grid  resistor  Ri. 

The  pi-network  circuit  of  figure  HC  is 
interesting  in  that  the  rotary  coil  L:  and  the 
plate  tuning  capacitor  G are  ganged  together 
by  a gear  train,  enabling  the  circuit  to  be  tuned 
to  resonance  with  one  panel  control  instead  of 
the  two  required  by  the  circuit  of  figure  llA. 
Careful  design  of  the  rotary  inductor  will  per- 
mit the  elimination  of  the  auxiliary  high  fre- 
quency coil  L],  reducing  the  cost  and  complexi- 
ty of  the  circuit. 

A single  tuned  circuit  may  be  employed  to 
couple  the  power  amplifier  grid  circuit  to  the 
buffer  stage  as  shown  in  figure  12.  Care  must 
be  taken  to  ensure  that  a good  ground  exists 
between  the  two  circuits. 

29-5  A Compact 

Linear  Amplifier  for 
Mobile  SSB 

Single  sideband  transmission  is  ideal  for 
mobile  operation  because  it  provides  the  max- 
imum amount  of  "talk  power”  for  a given 
amount  of  primary  power  drain.  The  new  three 
phase  alternator  systems  available  for  automo- 
tive use  are  capable  of  supplying  over  one 
thousand  watts  of  primary  power,  making  the 
proposition  of  a kilowatt  sideband  transmitter 
a reality. 


The  compaa  kilowatt  linear  amplifier  de- 
scribed in  this  section  is  designed  for  remote 
control  operation  on  any  one  amateur  band.  It 
may  be  driven  by  any  exciter  capable  of  one 
or  two  watts  p.e.p.  output. 

Amplifier  The  circuit  of  the  mobile  linear 
Circuif  amplifier  is  shown  in  figure  14. 

Designed  around  two  Eimac 
4W-300B  water  cooled  tetrode  tubes,  the  am- 
plifier may  nevertheless  be  used  with  4X-250B 
or  4CX-300A  tubes  if  air  cooling  is  employed. 
For  purely  mobile  operation  the  water  cooled 
tubes  are  recommended  as  air  blowers  are  bul- 
ky, noisy,  and  draw  a high  current  from  the 
automobile  primary  power  system.  Water  cool- 
ing, although  unfamiliar  to  many  amateurs,  is 
relatively  simple  and  inexpensive.  A Stewart- 
Warner  fuel  pump  is  used  to  pump  distilled 
water  from  a reservoir  through  polyethylene 
tubes  to  the  anodes  of  the  4W-300B  tetrode 
and  back  through  other  tubes  to  the  reservoir. 
This  container  is  a surplus  one  gallon  G.I. 
gasoline  can  (figure  17).  The  temperature  of 
the  water  remains  under  100°F.  even  when  the 
transmitter  is  operated  for  extended  periods  of 
time.  The  water  jackets  of  the  tubes  are  con- 
nected in  series  as  shown  in  the  drawing.  It 
has  been  found  by  experiment  that  it  is  not 
necessary  to  have  any  other  form  of  cooling 
medium  to  maintain  the  tubes  at  a safe  tem- 
perature. The  water  pump  draws  a current  of 
0.2  amperes  at  maximum  pumping  speed.  It 
is  mounted  in  a corner  of  the  turtle-back  of 
the  automobile,  along  with  the  G.I.  water  can. 
The  pump  motor  is  connected  so  that  it  starts 
as  soon  as  filament  voltage  is  applied  to  the 
two  4W-300B  tubes. 

The  two  tetrode  tubes  are  connected  in  par- 
allel with  a parallel  resonant  plate  circuit  and 
a simple  "pi-type”  grid  circuit.  The  plate  cir- 
cuit consists  of  a minature  Jennings  GSLA-120 
variable  vacuum  capacitor  (G)  in  parallel 
with  a silver  plated  copper  tubing  tank  coil. 
Plate  voltage  is  fed  to  the  tubes  through  a r-f 
choke  and  the  anodes  of  the  tubes  are  coupled 
to  the  tank  circuit  through  a 500  ggfd.,  5 KV 
ceramic  capacitor.  The  low  impedance  antenna 
system  is  directly  tapped  at  a low  impedance 
point  on  the  tank  coil.  A small  6 volt  auto- 
mobile headlamp  is  placed  in  series  with  the 
antenna  lead  to  provide  a simple  indicator  of 
r-f  current.  Antenna  transfer  is  accomplished 
by  means  of  a minature  Jennings  RB-I  s.p.d.t. 
vacuum  relay. 

Special  low  impedance  sockets  for  these 
tubes  are  supplied  by  the  Eimac  Co.  and  their 
use  is  recommended.  The  sockets  have  a built- 


HANDBOOK 


Mobile  SSB  Amplifier  621 


4W-300B/ 

4X-250B 


Ct — T20  fitifd.,  5KV  variable  vacuum  capacitor.  Jennings  GSLA’120 

Li — (20  mefers):  15  turns  4f18e.,  Ys"  diam.,  long.  Ad/usi  to  resonance  with  Ji 
shorted  and  tubes  in  sockets. 

Lj^(20  meters);  6 turns,  y4-inch  silver  plated  copper  tubing,  21/2^^  i‘d.,  2*'  long. 

Ad'iust  antenna  tap  for  proper  loading* 

RFC, — 2.5  mh.  National  R-lOO 

RFCi — 225  turns  #28  manganin,  F/s*'  f^ng,  in  series  with  National  R-IOO 

choke.  Or,  heavy-duty  Raypar  RL-102  choke  may  be  substituted  for  these  two 
items. 

RYi — SPOT  relay,  d.c.  coll  to  match  battery  voltage.  Jennings  RB-1  vacuum  relay 
employed 

Bi,  Bi,  B3 — Indicator  pilot  lamps.  Bt  should  have  500  ma.  rating;  Bt,  60  mo.  rating; 
and  B3  may  be  automobile  headlight  lamp  heavy  enough  to  carry  antenna 
current. 


in  screen  capacitor  which  provides  an  extreme- 
ly low  impedance  screen  ground  return  circuit, 
reducing  the  problem  of  parasitics  to  a mini- 
mum. 

A simple  "series”  tuned  input  circuit  is 
employed  in  the  grid  circuit.  The  internal  in- 
put capacitance  of  the  tubes  is  effectively  in 
parallel  with  the  grid  coil  resonating  it  to  the 
operating  frequency.  Since  no  grid  current  is 
drawn  during  class  ABl  operation  a small  bat- 
tery may  be  used  to  provide  the  correct  value 
of  operating  bias.  Flashlight  bulbs  are  placed 
in  the  screen  and  negative  high  voltage  leads 
instead  of  meters  to  conserve  space. 


Figure  15 

SMALL  SIZE  OF  LINEAR  AMPLIFIER 
IS  DUE  TO  USE  OF  MINIATURE 
300  WATT,  WATER  COOLED 
TETRODE  TUBES 

Parallel  operation  of  the  4W-300B  tubes  is 
employed.  Plate  tank  inductor  is  at  right, 
with  miniature  variable  vacuum  capacitor  be- 
hind it.  Plate  choke  is  in  the  foreground, 
made  up  of  a 2.5  mh.  choke  in  series  with  a 
home-made  unit.  Commercial  choke  may  be 
used,  as  indicated  in  parts  list  of  figure  14. 
Special  low  inductance  Cimac  sockets  are  em- 
ployed with  the  tubes.  Vacuum-type  antenna 
relay  is  at  tap,  center. 


Amplifier  Class  ABl  operation  plus  the 

Construction  complete  absence  of  parasitics 

results  in  a very  minimum 

amount  of  harmonic  generation.  The  use  of  a 
low  inductance  variable  vacuum  capacitor  pro- 
vides an  effective  return  path  to  ground  for 
high  frequency  harmonics  appearing  in  the 


622  H.F.  Power  Amplifiers 


THE  RADIO 


plate  circuit  of  the  amplifier.  As  a result,  a very 
minimum  of  shielding  and  lead  filtering  Is 
required  with  this  amplifier  even  when  28  Me. 
operation  is  desired.  Placing  the  amplifier 
within  the  turtle-back  of  the  automobile  and 
the  insertion  of  a low  pass  TV-filter  in  the 
coaxial  antenna  line  reduces  TVI-producing 
harmonics  to  such  a low  value  that  no  inter- 
ference is  caused  to  a nearby  television  receiver 
tuned  to  a channel  2 signal  of  moderate 
strength. 

The  amplifier  is  built  upon  a shallow  metal 
deck  just  large  enough  to  hold  the  various 
components  (figure  15).  The  two  4W-300B 
tube  sockets  are  placed  in  the  left  area  of  the 
chassis  with  the  variable  vacuum  capacitor 
mounted  at  the  right  corner  of  the  front  panel. 
The  silver  plated  copper  tubing  tank  coil  is 
bolted  between  the  rear  terminal  of  the  capaci- 
tor and  one  of  the  capacitor  mounting  bolts. 
Make  sure  that  a good  connection  exists  be- 
tween the  capacitor,  the  panel,  and  the  chassis 
as  this  ground  return  is  part  of  the  plate  tank 
circuit. 

Centered  on  the  front  panel  are  the  three 
pilot  lamp  indicator  sockets,  the  antenna 
changeover  relay  and  the  coaxial  antenna  re- 
ceptacles. At  the  rear  of  the  chassis  is  the  plate 
r-f  choke.  Two  r-f  chokes  in  series  were  used 
in  the  original  model,  but  have  since  been  re- 
placed with  one  of  the  new  Raypar  miniature 
chokes  (see  parts  list,  figure  14). 

The  grid  circuit  components  are  mounted 
in  the  under-chassis  area  of  the  amplifier  (fig- 
ure l6).  A shield  plate  is  placed  over  the 
bottom  of  the  amplifier  to  completely  isolate 
the  components  from  the  field  of  the  plate 
tank  circuit.  Grid  coil  Li  is  mounted  to  the 
panel  and  slug-tuned  to  resonance.  The  grid 
terminals  of  the  tube  sockets  are  connected 
with  a short  length  of  54 -inch  wide  copper 


Strap.  A wire  interconnecting  lead  will  invari- 
bly  lead  to  VHP  parasitic  oscillation  of  the 
stage. 

Amplifier  The  amplifier  must  be  completely 
Operation  tested  before  it  is  installed  in  the 
automobile.  Grid,  plate,  and  screen 
meters  should  be  inserted  in  the  power  leads 
to  the  amplifier  and  preliminary  tune-up  into 
a dummy  load  is  done  at  reduced  screen  and 
plate  potentials.  A small  driving  signal  is  ap- 
plied to  the  amplifier  and  grid  coil  Li  reson- 
ated. The  screen  and  plate  currents  will  in- 
crease sharply  in  value  when  the  grid  circuit 
is  tuned  to  the  operating  frequency.  The  degree 
of  plate  loading  is  determined  by  the  position 
of  the  antenna  tap  on  the  plate  tank  coil. 

Screen  current  is  a very  sensitive  indicator 
of  the  degree  of  amplifier  antenna  loading.  At 
plate  potentials  of  2000  or  so  the  total  screen 
current  at  one  kilowatt  peak  input  is  about  10 
milliamperes,  varying  slightly  with  individual 
tubes.  As  the  plate  potential  is  raised  above 
this  value  the  screen  current  gradually  drops 
until  at  3000  volts  plate  potential  the  screen 
current  is  near  zero.  In  general,  overcoupling 
of  the  amplifier  will  be  indicated  by  low 
screen  current  and  undercoupling  will  result 
in  high  values  of  screen  current.  At  3000  volts 
plate  potential,  an  overcoupled  condition  is 
indicated  by  negative  screen  current,  and  un- 
dercoupling will  produce  positive  screen  cur- 
rent. Only  a slight  deviation  from  zero  current 
during  voice  operation  indicates  proper  opera- 
tion. Maximum  peak  input  under  these  condi- 
tions is  about  1500  watts  (3000  volts  at  500 
ma.). 

The  Power  Supply  This  amplifier  is  designed 
to  work  from  a three- 
phase  alternator  system  of  1500  watts  capacity. 
A three-phase  rectifier  supply  is  employed  to 


Figure  16 

UNDER-CHASSIS  VIEW  OF  MOBILE 
SSB  LINEAR  AMPLIFIER  SHOWS 
SIMPLICITY  OF  CIRCUIT 

Input  circuit  is  at  the  left,  with  grid  terminals 
of  socket  connected  with  short  copper  strap. 
Screen  series  resistors  are  at  center,  with  grid 
choke  and  power  plug  at  right.  Under-chassis 
area  is  enclosed  with  aluminum  bottom  plate 
held  in  place  with  self-tapping  sheet  metal 
screws. 


HANDBOOK 


Mobile  SSB  Amplifier  623 


provide  2500  volts  at  400  milliamperes  (fig- 
ure 18).  Since  the  frequency  of  the  alternator 
is  120  cycles  or  higher  and  the  ripple  output 
of  the  supply  is  less  than  5%  before  filtering, 
a small  capacitor  is  sufficient  to  produce  pure 
d.c.  and  to  provide  good  regulation  for  peak 
power  surges. 

In  the  interest  of  power  conservation  and 
compactness  the  new  Sarkes-Tarzian  silicon 
rectifiers  are  used  in  the  high  and  low  voltage 
power  supplies.  These  rectifiers  can  be  operat- 
ed in  series  with  no  special  selection.  It  is 
possible  therefore  to  provide  compact  and  in- 
expensive high  voltage  rectifiers  by  connecting 
a number  of  the  standard  type  M-500  recti- 
fiers in  series  to  obtain  the  required  peak  in- 
verse voltage  rating.  The  higher  cost  of  the 
bank  of  silicon  rectifiers  as  compared  to  va- 
cuum tube  rectifiers  is  offset  by  the  elimina- 
tion of  the  tubes  and  a special  three-phase  fil- 
ament transformer.  If  desired,  the  special 
Sarkes-Tarzian  280-SM  rectifier  stack  may  be 
substituted  for  the  individual  units.  Since  the 


Figure  17 

WATER  CIRCULATION  SYSTEM 
FOR  X-531  (4W-300B)  TUBES 


alternator  will  not  supply  much  more  than 
1500  watts,  the  primary  voltage  drops  rapidly 


Ti  (SBC-  M 1 ) 


T 1 (sec.  #2 ) 


Figure  18 

SCHEMATIC,  THREE-PHASE  MOBILE  POWER  SUPPLY 

Di-D? — 500  mo.  silicon  rectifier.  Sarkes-Tarzian  M-500 

D7-Di2 — 2.5  KV,  730  ma.  rectifier  stack.  Sarkes-Tarzian  280-SM  silicon  rectifier.  See  text  for 
substitute 

Ti — Special  three-phase  power  transformer  to  deltYcr  2500  volts  d.c.  at  400  ma.,  and  325 
volts  d.c.  at  700  ma.  Design  data  and  core  may  be  obtained  from  Arnold  Engineering  Co., 
Marengo,  III.  Core  number:  ATA-1573  with  12-mil  Hypersil  laminations  for  720  cvcie 
operation. 

Primary  winding:  (12  volts) — Each  leg,  IS  turns  ifSe.  wire 
( 6 volts) — Each  leg,  7 turns  #5e.  wire 
Low  Voltage  secondary.  Each  leg,  265  turns  #30e.  wire 
High  Voltage  secondary:  Each  leg,  2400  turns  #28e.  wire 

7.5  KV  insulation  employed.  High  voltage  winding  placed  next  to  core,  low  voltage 
secondary  next,  and  primary  winding  outside,  with  2KV  insulation.  Vacuum  impregnate 
windings.  When  finished,  po7orize  windings  by  placing  6 volt,  50  c.p.  lamps  in  series 
with  each  primary  lead.  Bulbs  should  only  glow  red  when  primary  windings  are  correctly 
polarized.  Correct  polarization  of  secondary  windiruis  will  deliver  desired  output  voltaaes. 


under  overload  conditions  and  the  silicon  rec- 
tifiers cannot  be  damaged,  even  with  a direct 
short  circuit  across  the  output  terminals  of  the 
high  voltage  power  supply. 

This  unusual  equipment  is  described  for 
others  who  may  wish  to  make  models  based 
upon  ideas  presented  in  the  design.  Although 
some  of  the  components  shown  in  the  ampli- 
fier or  power  package  are  unique  or  are  not 
readily  available,  it  is  hoped  that  the  reader 
can  make  use  of  his  facilities  and  abilities  to 
use  many  of  the  ideas  shown  herewith. 

Variations  Upon  It  is  possible  to  increase 

the  Basie  Circuit  the  power  level  of  this 

amplifier  by  merely  add- 
ing more  tubes  connected  in  parallel  with  the 
original  two.  Three  tubes  will  permit  a p.e.p. 
of  two  kilowatts  at  3000  volts  plate  potential. 


Figure  1 9 

WATER  COOLED  LINEAR 
AMPLIFIER  PACKAGED  FOR 
FIXED  STATION  USE 

Either  4W-300B  (water  cooled)  tubes  or  4X- 
250B  (air  cooled)  tubes  may  be  employed  in 
this  compact  linear  amplifier.  Input  and  out- 
put capacitors  of  pi-network  ore  vacuum 
units,  and  power  supply  (three  phase)  is  con- 
tained within  the  cabinet  (right).  Wide  band 
plate  r.f.  choke  is  at  left  of  cabinet,  running 
from  front  to  back.  Amplifier  operates  at 
plate  potential  of  3000  volts,  with  p.e.p,  of 
1500  watts. 

Four  tubes  in  parallel  will  provide  a p.e.p.  of 
two  kilowatts  at  2000  volts  plate  potential.  An 
experimental  amplifier  having  five  tubes  in 
parallel  has  run  3500  watts  p.e.p.  into  dummy 
loads.  The  five  tube  amplifier  employed  neu- 
tralization and  parasitic  suppression  and  per- 
formed in  normal  fashion  with  no  signs  of  in- 
stability. 

Shown  in  figure  19  is  a fixed  station  version 
of  this  amplifier.  Employing  a pi-section  plate 
tank  circuit  and  a three-phase  power  supply, 
the  complete  unit  is  built  within  a standard 
12"  X 16"  X 19"  cabinet.  Of  unusual  interest 
is  the  r-f  choke  employed  in  the  plate  circuit 
of  the  linear  amplifier  stage.  This  choke  is 
usually  a critical  item  as  the  full  value  of  r-f 
plate  potential  is  developed  across  it.  For  max- 
imum amplifier  efficiency  and  minimum  loss 
the  plate  choke  must  present  an  impedance  of 
several  thousand  ohms  at  any  frequency  of 
operation.  Special  r-f  chokes  are  available  that 
will  perform  this  exacting  task,  and  the  one 
described  equals  the  best  of  the  commercial 
items  at  a fraction  of  the  cost.  The  choke  coil 
consists  of  250  turns  of  #22  double  cotton 
covered  manganin  wire  close-wound  upon  a 
14-inch  diameter  wooden  dowel  rod.  The  small 
d.c.  resistance  of  the  winding  effectively  de- 
stroys undesired  resonance  effects  within  the 
choke,  providing  a high  and  uniform  imped- 
ance across  ail  amateur  bands  from  80-  to  10- 
meters.  The  choke  will  work  well  up  to  power 
levels  of  the  order  of  2000  watts,  p.e.p.,  or 
1000  watts,  plate  modulated. 

29-6  A Mulfi-band 

Mobile  Linear  Amplifier 


Figure  20 

MULTI-BAND  LINEAR  AMPLIFIER 
FOR  MOBILE  SERVICE  FEATURES 
THREE  SEPARATE  TANK  CIRCUITS 
SWITCHED  BY  CONTROL  RELAYS 

Three  vacuum  tank  capacitors  are  located  at 
top  left  corner  of  panel.  To  the  right  is  ther- 
mocouple r.f.  ammeter  in  output  circuit.  Zero- 
center  screen  meter  and  0-500  d.c.  ma.  plate 
meter  are  at  lower  right. 


The  kilowatt  linear  amplifier  described  in 
this  section  is  designed  for  remote  control  mo- 
bile operation  on  any  three  amateur  bands. 
The  unit  shown  in  the  photographs  operates 
on  the  80,  40,  and  20  meter  bands,  but  opera- 
tion on  15  or  10  meters  is  possible  by  altering 
the  constants  of  the  plate  tank  circuit. 


625 


Figure  21 

SCHEMATIC,  MULTI-BAND  MOBILE  LINEAR  AMPLIFIER 

Ct,  Ci,  Cs^Jennings  variable  vacuum  capacitor,  type  GSLA.  Use  250  li/ifd.  unit  for  BO  meters  and  120 
ti^lfd.  units  for  other  bands.  Adjust  coils  to  resonate  circuit  with  capacitors  set  near  maximum  value. 
Li — 76  turns,  i/t-inch  silver  plated  copper  tubing,  2I/2"  long.  Antenna  tap  approximately  2 

turns  from  ground  end. 

8 turns,  same  as  Li.  Antenna  tap  approximately  2 turns  from  ground  end. 

Lj — 5^/2  turns,  same  as  Li.  Antenna  tap  approximately  ^ turn  from  ground  end. 

RFCi — 2.5  mh.  National  R-700 

RFCi — Wide  band  choke,  500  ma.  rating.  Ray  par  RL~102 

RYi,  i,  3—^OPDT  relay,  d.c.  coil  to  match  battery  voltage.  Jennings  type  RB  vacuum  relay 
RYi — SPST  relay,  d.c.  coil  to  match  battery  voltage.  Jennings  type  RB-1  vacuum  relay  (Jennings  Radio 
Co.,  San  Jose,  Calif.) 

Ml — 0 - 500  mo.  d.c.  meter 

M2 — 500  - 0 - 500  d.c.  micro-ammeter,  zero  center  movement,  shunted  to  50  - 0 - SO  ma. 

Mj — 0 - 5 ompere,  r.f.  ammeter 


Amplifier  The  amplifier  employs  two  ex- 
Circuit  ternal  anode  tetrodes  such  as  the 

4W-300B,  4X-250B,  or  4CX- 
300A.  The  water  cooled  tubes  are  recommend- 
ed for  mobile  operation,  while  the  air  cooled 
tubes  may  be  used  for  fixed  operation.  The 
schematic  of  the  amplifier  is  shown  in  figure 
21  and  is  a version  of  the  untuned  input  cir- 
cuit of  figure  IIC. 

Three  separate  parallel  tuned  plate  circuits 
are  employed  in  the  linear  amplifier.  Each 
circuit  is  made  up  of  a silver  plated  copper 
tubing  coil  paralleled  with  a miniature  vari- 
able vacuum  tuning  capacitor.  These  new  Jen- 
nings capacitors  are  extremely  small  in  size 
and  are  tuned  by  means  of  a slotted  shaft  in- 
stead of  the  more  cumbersome  and  expensive 
counter  dial  arrangement.  The  low  impedance 
mobile  antenna  system  is  tapped  directly  to 
experimentally  determined  points  on  the  tank 


coils.  The  proper  tuned  circuit  is  switched  to 
the  linear  amplifier  by  means  of  miniature 
vacuum-sealed  d.p.d.t.  relays.  Three  relays  are 
required,  one  for  each  tank  circuit.  The  tank 
circuits  are  relatively  low  impedance  devices 
as  they  are  tuned  with  capacitors  having  a large 
value  of  maximum  capacitance,  permitting 
good  efficiency  in  matching  the  tubes  to  the 
very  low  load  impedance  presented  by  the 
mobile  antenna  system. 

Although  the  tubes  operate  with  "zero” 
driving  power  the  resistive  input  system  de- 
mands that  approximately  9 watts  of  excita- 
tion be  employed  to  develop  50  volts  (peak) 
across  the  300  ohm  circuit  impedance.  The 
mobile  SSB  exciters  of  Chapter  28  used  in 
conjunction  with  a 2E26  linear  buffer  stage 
will  supply  an  abundance  of  excitation  for 
this  kilowatt  amplifier.  Alternatively,  a high 
impedance  grid  circuit  such  as  shown  in  figure 


626  H.F.  Power  Amplifiers 


THE  RADIO 


1 IB  may  be  used,  dropping  the  grid  driving 
power  level  to  a watt  or  so. 

Selection  of  the  proper  tank  circuit  is  ac- 
complished by  means  of  a rotary  switch  (S2) 
located  at  the  driver’s  position  in  the  automo- 
bile. One  of  the  three  miniature  plate  circuit 
relays  at  a time  may  be  operated  by  this  switch. 
The  antenna  change-over  relay  is  actuated  by 
the  push-to-talk  circuit  in  the  microphone. 

Amplifier  The  amplifier  is  built  upon 

ConsfrucHon  a shallow  metal  deck  (figure 
22 ) . The  two  Eimac  low  in- 
ductance sockets  for  the  tetrode  tubes  are 
placed  in  the  rear  corner  of  the  chassis  with 
the  80  meter  tank  coll  to  their  left.  One  end 
of  the  coil  is  bolted  to  the  chassis  and  the 
other  end  is  attached  to  the  rear  terminal  of 
the  variable  vacuum  capacitor  which  is  mount- 
ed to  the  metal  panel.  The  40  and  20  meter 
tank  coils  are  mounted  directly  to  their  respec- 
tive tuning  capvacitors. 

The  front  area  of  the  chassis  is  taken  up 
with  the  plate  circuit  switching  relays  RYi-2-.'i. 
These  relays  are  bolted  to  the  chassis  and 
connections  are  made  to  the  various  terminals 
by  means  of  flexible  1/4 -inch  silver  plated  cop- 
per strap.  The  three  meters  are  mounted  in  the 
remaining  panel  space.  Under-chassis  wiring 
is  extremely  simple,  as  seen  in  figure  23.  The 
socket  grid  terminals  are  connected  together 
with  a short  length  of  copper  strap.  A shield 
is  not  required  over  the  bottom  of  the  chassis 
since  the  grid  circuit  is  untuned. 


Amplifier  The  amplifier  should  be  bench- 
Operation  tested  before  it  is  placed  within 
the  automobile.  Operating  con- 
ditions are  similar  to  those  described  in  Section 
29-5.  A center-reading  milliammeter  is  used 
in  the  screen  circuit  to  observe  screen  current 
during  loading  adjustments.  Linear  operation 
may  be  checked  with  the  aid  of  envelope  de- 
tectors, as  discussed  in  an  earlier  chapter.  Nor- 
mal operation  takes  place  at  2500  volts  at  400 
milliamperes,  peak  current.  Screen  current  is 
less  than  5 milliamperes  under  these  condi- 
tions. Grid  current  is  zero. 

29-7  An  Inexpensive 
Cathode  Driven 
Kilowatt  Amplifier 

An  objectionable  feature  of  triode  tubes  is 
that  they  must  be  neutralized  in  conventional 
grid  driven  circuits.  Tetrode  tubes  may  often 
dispense  with  the  neutralizing  circuit,  but  they 
require  a screen  power  supply  whose  cost  and 
complexity  dissipate  the  saving  afforded  by 
the  elimination  of  the  neutralizing  circuit.  The 
use  of  a cathode  driven  amplifier  employing 
zero  bias  tubes  overcomes  these  two  disadvan- 
tages. The  grids  act  as  an  excellent  screen  be- 
tween the  plate  and  cathode  so  neutralization 
is  not  usually  required.  The  very  small  plate 
to  cathode  capacitance  permits  a minimum  of 
intercoupling  on  frequencies  below  30  Me. 
when  inexpensive  tetrode  or  pentode  tubes  are 
used.  No  screen  or  bias  supplies  are  required. 


Figure  22 
REAR  VIEW  OF 
MULTI-BAND 
MOBILE  LINEAR 
AMPLIFIER 

The  three  amplifier  tank 
coils  are  connected  be- 
tween the  variable  va- 
cuum capacitors  and 
chassis  ground.  Plate 
voltage  is  shunt-fed  to 
the  linear  amplifier  tubes 
(4W-300B's  or  4X- 
2S0B's}.  Antenna  reiay 
is  at  left.  Extremely  low 
harmonic  content  of  class 
AB 1 amplifier  operation 
requires  no  TV!  screen- 
ing or  other  enclosure  to 
prevent  radiation  of  un- 
desired harmonics. 


HANDBOOK 


Multi-Band  Amplifier  627 


Feedthrough  A large  portion  of  the  exciting 
Power  power  appears  in  the  plate 

circuit  of  the  cathode  driven 
stage  and  is  termed  the  feedthrough  power. 
In  any  amplifier  of  this  type,  whether  it  be 
triode  or  tetrode,  it  is  desirable  to  have  a fairly 
large  ratio  of  feedthrough  power  to  peak  grid 
driving  power.  The  feedthrough  power  acts  as 
a swamping  resistor  across  the  driving  circuit 
to  reduce  the  effect  of  grid  loading.  The  ratio 
of  feedthrough  power  to  driving  power  should 
be  at  least  10  to  1 for  best  stage  linearity. 

Class  AB1  Cathode  The  only  way  that  the 
Driven  Amplifiers  advantages  can  be  taken 
of  the  inherent  feed- 
back in  a cathode  driven  amplifier  is  to  drive 
it  with  a source  having  a high  internal  im- 
pedance and  to  operate  the  stage  class  ABl 
without  grid  or  screen  current.  It  is  a char- 
acteristic of  class  ABl  cathode  driven  (so- 
called  "grounded  grid")  amplifiers  that  a vari- 
ation in  load  impedance  on  the  stage  reflects 
back  as  a nearly  proportional  change  at  the 
input  because  the  input  and  output  circuits 
ate  essentially  in  series.  In  the  same  manner,  a 
change  in  gain  (such  as  may  be  caused  by  a 
nonlinearity  of  gain  with  respect  to  signal 
level)  of  a cathode  driven  stage  reflects  back 
as  a change  in  the  input  impedance  of  that 
stage.  If  the  cathode  driven  stage  is  fed  by  a 
source  having  high  internal  impedance  such 
as  a tetrode  amplifier,  changes  in  gain  or  non- 
linearity in  the  cathode  driven  stage  will  be 
compensated  for.  For  example,  if  the  gain  of 
the  cathode  driven  stage  drops  slightly,  the  in- 
put impedance  of  the  stage  will  increase  and 
the  load  impedance  on  the  driver  stage  will 
rise.  If  the  driver  has  high  internal  impedance 
the  output  voltage  will  also  rise  when  its  load 
impedance  rises.  The  increased  output  voltage 
will  raise  the  output  voltage  of  the  cathode 
driven  stage  so  that  the  output  is  nearly  up  to 
where  it  should  be  even  though  the  gain  of 
the  cathode  driven  stage  is  low.  The  equivalent 
circuit  for  this  action  is  shown  in  figure  24. 

"True”  class  ABl  cathode  driven  amplifiers, 
however,  have  rather  low  stage  gain  and  the 
use  of  low-mu  triodes  is  required  for  reason- 
able efficiency.  Such  low  gain  circuits  require 
more  stages  to  reach  a given  power  level.  High 
driving  voltage  is  required  which  is  difficult 
to  achieve  in  the  cathode  circuit.  Since  i'l  s'tems 
impractical  to  operate  most  available  tubes 
class  ABl  in  a cathode  driven  circuit  (i.e.,  no 
grid  current  and  no  screen  current)  the  possi- 
bility of  taking  advantage  of  the  inherent  feed- 
back in  the  circuit  is  lost  because  a low  driver 


Figure  23 

UNDER-CHASSIS  VIEW  OF  MULTI- 
BAND LINEAR  AMPLIFIER 

Coaxial  input  receptacle  is  at  left,  with  com- 
position-type grid  resistors.  Simplicity  of  cir- 
cuit requires  relatively  few  components  be- 
neath the  chassis.  Grid  terminals  of  sockets 
are  connected  with  wide  copper  strap. 


source  impedance  is  then  required  to  supply 
grid  driving  power  for  class  AB2  operation. 
The  latter  class  of  operation  offers  stage  gain 
figures  of  10  to  25  with  moderate  values  of 
cathode  excitation  voltage.  The  linearity,  how- 
ever, is  approximately  the  same  as  for  con- 
ventional grid  driven  circuits  having  low 
source  impedance,  and  about  30  decibels  sig- 
nal-to-distortion  ratio  is  the  maximum  that 
can  be  achieved. 


Vi  Va  V3 


® 

Figure  24 

EQUIVALENT  CIRCUIT  OF  CASCADE 
CLASS  ABl  CATHODE  DRIVEN 
AMPLIFIERS 

A — Cathode  driven  amplifiers  may  be  ar- 
ranged in  cascade  to  provide  high  output 
levels.  Any  variation  in  load  impedance 
(R)  reflects  back  as  nearly  proportional 
change  at  the  input,  as  input  and  output 
circuits  are  essentially  in  series  as  shown 
in  figure  B. 

B — Equivalent  circuit  of  figure  A.  Ri  has  a 
very  high  resistance,  and  Rz  and  Rs  have 
very  low  resistances  compared  to  the 
load  R. 


628  H.F.  Power  Amplifiers 


Figure  25 

CATHODE  DRIVEN  KILOWATT 
AMPLIFIER  FOR  40  AND  80  METERS 

Cathode  driven  amplifier  presents  ultimate  in 
simplicity  and  stability.  Designed  around  two 
803  tetrodes,  this  kilowatt  amplifier  for  40 
and  60  meter  operation  has  only  plate  tank 
tuning  control  and  plate  current  meter  on 
panel.  Amplifier  is  completely  enclosed  for 
reduction  of  spurious  radiation. 


A Practical  Cathode  Shown  in  figures  25, 

Driven  Amplifier  27,  and  28  is  a cathode 

driven  amplifier  de- 
signed for  40  and  80  meter  operation.  It  is 
capable  of  1000  watts  power  input  and  may 
be  employed  for  SSB,  c.w.,  or  a.m.  phone. 
Combined  grid  driving  and  feedthrough  power 
is  approximately  110  watts  for  1 kilowatt  in- 
put. Actual  grid  driving  power  is  of  the  order 
of  9 to  10  watts,  so  the  ratio  of  feedthrough  to 
drive  power  is  about  10  to  1. 

The  schematic  of  the  amplifier  is  shown  in 
figure  26.  Two  803  tetrode  tubes  are  used  in 
parallel  in  cathode  driven  configuration.  These 
tubes  are  available  on  the  surplus  market  for  a 
dollar  or  so  and  perform  very  well  up  to  ap- 
proximately 10  Me.  All  three  grids  of  the 
tubes  are  returned  to  ground  and  the  tubes  op- 
erate in  near  zero  bias  condition.  The  internal 
screening  of  the  tubes  begins  to  deteriorate 
in  the  15  Me.  region  and  amplifier  instability 
may  be  encountered  above  10  Me.  unless  pre- 
cautions are  taken  to  ensure  that  the  grids  re- 
main close  to  r-f  ground  potential. 

When  operating  in  the  3.5-7  Me.  region 
it  is  entirely  possible  to  apply  the  r-f  driving 
voltage  directly  to  the  filament  circuit  without 
the  necessity  of  adding  filament  r-f  chokes  to 
isolate  the  driving  voltage  from  the  filament 
transformer.  This  makes  for  circuit  simplicity 
and  lower  initial  cost.  If  this  circuit  is  used 


with  high  frequency  tubes  such  as  the  813 
and  7094,  the  use  of  filament  chokes  for  op- 
eration above  7 Me.  is  recommended. 

This  type  of  amplifier  has  a tendency  to- 
ward parasitic  oscillations  in  the  100  Me.  re- 
gion at  which  point  the  inherent  inductance  of 
the  ground  paths  begins  to  affect  operation. 
Variations  in  grounding  technique  can  elimin- 
ate the  parasitic,  but  the  addition  of  parasitic 
suppressors  in  the  plate  leads  of  the  tubes 
makes  the  circuit  relatively  insensitive  to  vari- 
ations in  the  ground  return  path. 

Amplifier  The  amplifier  is  built  upon  a 

ConsfrucKon  10"  x 19"  x 4"  aluminum 
chassis.  The  plate  tuning  ca- 
pacitor Cl  is  centered  on  the  chassis  with  the 
two  803  tubes  mounted  to  one  side.  The  tube 
sockets  are  recessed  below  the  chassis  to  bring 
the  top  of  the  tube  base  level  with  the  top 
of  the  chassis.  The  plate  r-f  choke  and  bypass 
capacitors  are  mounted  between  the  tubes.  On 
the  opposite  side  of  the  tank  capacitor  is  the 
tank  coil  and  the  r-f  thermocouple  (figure  27). 

In  the  under-chassis  area  the  only  compon- 
ents are  the  cathode  current  meter,  the  filament 
transformer,  and  the  small  parts  associated 
with  the  filament  circuit.  Pins  2,  3,  and  4 of 
each  socket  are  grounded  to  the  metal  rods 


1 1 5 V. 


Figure  26 

SCHEMATIC,  CATHODE  DRIVEN 
AMPLIFIER 

Cl — 2S0  iififd.,  5000  volt  spacing 
Li — 40  or  80  meter  inductor.  Cut  to  resonate 
to  frequency  with  Ci  near  full  capacity. 
Antenna  coil  5 turns. 

PC — 40  ohm,  2 watt  composition  resistor 
wound  with  6 turns  # 18  e.  wire 
TC — R.f.  thermocouple,  0-5  amperes 
Ti — 70  volts  at  10  amperes.  Stancor  P-6461 
RFCi — 1.0  mh.  Johnson  102-752 
RFC, — Heavy  duty  plate  choke.  National  R- 
7754 

Ml — 0-500  d.c.  miHiamperes 

M: — 0-5  amperes,  r.f.  (Used  with  TC) 


Figure  27 

REAR  VIEW,  CATHODE  DRIVEN 
803  AMPLIFIER 

Plate  tuning  capacitor  is  at  the  center^  with 
tank  coil  to  left.  The  803  tubes  (right)  are 
mounted  with  sockets  below  deck,  so  that 
tops  of  bases  are  level  with  chassis.  Plate 
r.f.  choke  is  placed  between  tubes,  with  para- 
sitic chokes  (PC)  supported  from  top  termi- 
nal of  choke. 


supporting  the  sockets.  The  complete  ampli- 
fier is  enclosed  in  a shield  made  of  aluminum. 
Holes  are  drilled  above  the  803  tubes  in  the 
top  sheet  (figure  25)  and  in  the  rear  plate  to 
permit  circulation  of  air  about  the  tubes. 

Amplifier  The  resting  plate  current  of  the 
Operation  amplifier  under  quiescent  condi- 
tion is  between  30  ma.  and  75 
ma.  depending  upon  the  plate  voltage,  climb- 
ing to  450  ma.  or  so  under  full  excitation  at  a 
plate  potential  of  2500  volts.  Care  should  be 
taken  in  tune-up  and  full  excitation  should 
neper  be  applied  unless  the  amplifier  is  op- 
erating at  rated  plate  voltage  with  maximum 
antenna  loading.  Applying  full  drive  to  the 
tubes  with  no  plate  voltage  will  surely  cause 
damage  to  the  grid  structure  of  the  tubes. 

The  metet  indicates  plate  and  screen  cur- 
rent; actual  plate  current  is  about  40  milliam- 
peres  lower  than  the  metet  reading.  At  1000 
watts  p.e.p.,  the  power  output  measures  ap- 
proximately 600  watts.  Tunin'^  of  the  plate 
circuit  is  facilitated  by  means  of  a thermo- 
couple t-f  meter  placed  in  the  coaxial  lead  to 
the  antenna  coaxial  receptacle. 


Figure  28 

UNDER-CHASSIS  VIEW  OF  CATHODE 
DRIVEN  AMPLIFIER 

input  coaxial  connector  is  at  the  right  on 

rear  of  chassis.  Mica  filament  bypass  capaci- 
tors are  placed  between  the  tube  sockets, 
with  RFCi  beneath  them. 


29-8  A Low  Distortion 
Sideband  Linear  Amplifier 

A properly  designed  grounded  grid  linear 
amplifier  is  a very  satisfactory  high  level  stage 
for  a sideband  transmitter,  even  though  the 
possibility  of  taking  advantage  of  the  inherent 
feedback  is  lost  because  of  the  reasons  explain- 
ed in  Section  29-7  of  this  chapter.  With  care, 
a signal-to-distortion  ratio  of  30  decibels  can 
easily  be  achieved  and  the  circuit  is  unsur- 
passed for  stability  and  ease  of  operation. 

Data  given  in  the  tube  manuals  and  instruc- 
tion sheets  are  usually  derived  for  maximum 
tube  output  rather  than  the  condition  of  min- 
imum distortion.  It  is  therefore  necessary  to 
adjust  circuit  operating  voltages  to  arrive  at  a 
satisfactory  power  output  level  consistent  with 
a pre-set  value  of  maximum  allowable  distor- 
tion. The  maximum  distortion  level  for  this 
amplifier  was  set  at  30  db  S/D.  At  this  figure 
350  watts  output  p.e.p.  can  be  obtained  at  3000 
volts  plate  potential,  and  680  watts  output 
p.e.p.  can  be  obtained  at  3500  volts. 

The  Grounded  For  maximum  shielding  with- 
Screen  in  the  tube,  it  is  necessary  to 

Configuration  operate  both  grid  and  screen 
at  r-f  ground  potential.  At  the 
same  time,  the  control  grid  has  a negative  op- 
erating bias  applied  to  it,  and  the  screen  has  a 
positive  operating  potential.  Both  elements  of 
the  tube,  therefore,  must  be  bypassed  to 
ground  over  a frequency  range  that  includes 
the  operating  spectrum  and  the  region  of  pos- 
sible VHF  parasitic  oscillations.  This  is  quite 
a large  order.  The  inherent  inductance  of  the 
usual  bypass  capacitors  is  sufficient  to  intro- 
duce sufficient  regeneration  into  the  circuit  to 
degrade  the  linearity  of  the  amplifier  at  high 
signal  levels  even  though  the  instability  is  not 
gteat  enough  to  cause  parasitic  oscillation.  The 


630  H.F.  Power  Amplifiers 


THE  RADIO 


© 


CONFIGURATION  PROVIDES  HIGH 
ORDER  OF  ISOLATION  IN  TETRODE 
AMPLIFIER  STAGE 


A — Typical  amplifier  circuit  has  cathode  re- 
turn at  ground  potential.  All  circuits  re- 
turn to  cathode. 

B — All  circuits  return  to  cathode,  but  ground 
point  has  been  shifted  to  screen  terminal 
of  tube.  Operation  of  the  circuit  remains 
the  same,  as  potential  differences  between 
elements  of  the  tube  are  the  same  as  in 
circuit  A. 

C — Practical  grounded  screen  circuit.  "Com- 
mon minus"  lead  returns  to  negative  of 
plate  supply,  which  cannot  be  grounded. 
Switch  removes  screen  voltage  for  tune- 
up  purposes. 


solution  to  this  problem  is  to  eliminate  the 
screen  bypass  capacitors  by  actually  grounding 
the  screen  terminals  of  the  amplifier  to  the 
chassis  by  means  of  a low  inductance  strap.  It 
is  possible  to  do  this  by  juggling  the  ground 
return  of  the  amplifier,  as  shown  in  figure  29. 
Circuit  A shows  the  conventional  d.c.  return 
paths  wherein  all  power  supplies  are  returned 
to  the  grounded  cathode  of  the  stage.  Meters 
are  placed  in  the  common  return  leads,  and 
each  meter  reads  only  the  current  flowing  in 
the  singular  circuit.  The  d.c.  ground  lead  is 
removed  from  the  cathode  in  circuit  B and 


placed  at  the  screen  terminal  of  the  tube.  Cir- 
cuit operation  is  still  the  same,  as  all  power 
supply  returns  are  still  made  to  the  cathode  of 
the  tube.  The  only  change  is  that  the  screen 
r.f.  ground  and  d.c.  ground  are  now  one  and 
rhe  same.  The  cathode  circuit  is  now  at  a 
negative  potential  with  respect  to  the  chassis 
ground  by  an  amount  equal  to  the  screen 
voltage.  In  addition,  the  return  of  the  high 
voltage  plate  supply  is  now  negative  with  re- 
spect to  ground  by  an  amount  equal  to  the 
screen  voltage  supply. 

A practical  version  of  this  circuit  is  shown 
in  figure  29C.  The  grid  bias  and  screen  sup- 
plies are  incorporated  in  the  amplifier  and 
separate  terminals  are  provided  for  the  posi- 
tive and  negative  leads  of  the  high  voltage 
supply.  A "tune-operate”  switch  5=  is  added  to 
the  circuit  which  removes  the  screen  potential 
for  tune-up  purposes.  The  negative  of  the 
high  voltage  supply  "floats  below  ground”  at 
the  value  of  the  screen  voltage. 

This  circuit  is  unique  in  that  it  requires  a 
mental  orientation  of  the  user  to  accept  the  fact 
that  d.c.  ground  is  no  longer  "ground.”  How- 
ever, the  circuit  is  no  more  complex  than  the 
usual  grounded  cathode  d.c.  return  circuit,  and 
the  advantage  of  having  the  screen  at  r.f. 
ground  potential  far  outweighs  the  unusual 
aspects  of  the  circuit.  Operation  of  the  circuit 
is  normal  in  ail  respects,  and  it  may  be  applied 
to  any  form  of  tetrode  amplifier  with  good 
results. 

The  Amplifier  The  circuit  of  an  operation- 
Circuif  al  cathode  driven,  grounded 

screen  amplifier  is  shown  in 
figure  30.  A 4-2 50A  or  4-400 A tetrode  tube 
is  used  with  an  untuned  cathode  circuit  and  a 
pi-network  plate  circuit.  Bias  and  screen  sup- 
plies are  included  in  the  amplifier.  The  im- 
portant operating  parameters  are  also  shown 
in  the  diagram.  P.e.p.  excitation  is  12  watts 
(including  losses)  for  3000  volt  operation 
and  about  22  watts  (including  losses)  for 
3500  volt  operation. 

The  resonant  plate  circuit  is  capaoitively 
coupled  to  the  amplifier  tube  and  consists  of 
a 300  ggfd.  variable  vacuum  capacitor,  a kilo- 
watt tapped-coil  turret,  and  a 1500  ggfd.  var- 
iable air-type  loading  capacitor.  D.c.  ground 
is  established  by  returning  the  output  section 
of  the  network  to  ground  by  means  of  choke 
RFC-3.  Three  .001  gfd.,  5 KV  ceramic  capaci- 
tors in  parallel  form  the  plate  blocking  capaci- 
tor. The  high  voltage  is  applied  to  the  tube 
via  a Raypar  heavy  duty  r.f.  choke  in  series 
with  two  air  wound  VHF  choke  coils.  Escape 


HANDBOOK 


SSB  Linear  Amplifier  631 


CLASS  ABl  OPERATI  NG  CONDITIC 
4-  250A  / 4-400A 


D.c.  PLATE  VOLTAGE 

3000 

3500 

O.C,  SCREEN  VOLTAGE 

300 

300 

STATIC  PLATE  CURRENT  (M^.) 

60 

60 

D.C.  GRID  BIAS  VOLTAGE 

'60 

-60 

MAX.SlG.  GRID  CURRENT 

3.6 

,0 

MAX.  SIG.  SCREEN  CURRENT  (MA) 

4,6 

MAX.  SlG.  PLATE  CURRENT  fM/4,) 

195 

£65 

MAX,  SIG.  PLATE  DISSIP.  (WArrS) 

£35 

235 

STATIC  PLATE  DISSI  PATI0N(*V/4rrS) 

160 

£10 

GRID  DRIVING  POWER  (W/trrS) 

0.63 

3.4 

FEED  THRU  POWER  (W^rrs) 

6.6 

15.6 

POWER  OUTPUT  (w^rrs) 

350 

690 

EFFICIENCY  {%») 

60 

70 

C7 

RELAY 

CONTROL 


Figure  30 

SCHEMATIC,  LOW  DISTORTION  LINEAR  AMPLIFIER 

Cl — 300  maW-,  JOKV  variable  vacuum  capacitor.  Jennings  UCS-300 
C’ — 7500  variable  capacitor.  Cardwell  8073 

Cj — 700  fxfifd.  variable  ceramic 

CirCr — 0.7  fifd.,  600  volt  bulkhead-type  capacitor.  Sprague  ^'Hypass" 

Li-Si — Barker  & Williamson  #850  pi-network  inductor.  Inductances:  80  meters,  73.5 
jxh.;  40  meters,  6.5  fih.;  20  meters,  7.75  fib.;  75  meters,  7 /uh.;  70  meters, 

0.8  iih. 

L: — Dual  winding  filament  choke.  B&W  FC-15,  75  amperes  capacity 

Ti — 5 volts  at  70.5  amperes.  Stancor  P'6735 

T: — 335  - 0 - 335  vo/ts,  75  ma.  Chicago  PSC-70 

CHi — 8 henry,  85  mo.  Stoncor  C-I709 

Di-Di — 500  ma.  silicon  rectifier.  Sarkes-Tarzian  M-500 

D.5 — Diode,  type  7N67  or  equivalent 

Ml — 0 - 800  mo.  d.c.  meter.  Marion  type  MM3 

Mi — 0 - 50  mo.  d.c.  meter.  Marion  type  MM3 

Ms — 0 - 50  mo.  d.c.  meter.  Mon'or?  type  MM3 

RFCi — Heavy  duty  wide  band  r.f.  choke,  600  mo.  Raypar  RL-100 

RFC2 — 20  turns  #20e,  airwound,  1/2"  diam,  V'  long 

RFCs — 2.5  mh.  National  R-100 


of  r.f.  energy  through  this  lead  is  thus  held 
to  an  absolute  minimum  value. 

Driving  power  is  coupled  to  the  filament 
circuit  of  the  tube  through  two  .001  /rfd.  cer- 
amic capacitors.  The  filament  transformer  is 
Isolated  from  the  exciting  voltage  by  means  of 
a bifilar  wound  r.f.  choke  coil,  capable  of  car- 
rying the  filament  current  of  the  tube  (10.5 
amperes ) . 

A dual  purpose  power  supply  delivering 
375  volts  serves  as  screen  and  bias  source.  The 
output  voltage  of  the  supply  is  held  constant 
by  three  regulator  tubes  in  series.  The  "com- 
mon minus”  lead  (see  figures  29  and  30)  is 
at  the  juncture  of  the  regulated  -75  volts  and 
the  regulated  300  volts.  The  supply  thus  de- 
livers -75  volts  for  the  bias  circuit  and  plus 
300  volts  for  the  screen  circuit.  The  exact  value 


Figure  31 

KHOWATT  CATHODE-DRIVEN 

AMPLIFIER 

Careful  arrangement  of  components  permits 
symmetrical  panel  layout.  Grid,  screen,  and 
plate  current  meters  are  spaced  across  top 
of  panel,  with  amplifier  loading  control  (Ct) 
at  left,  and  plate  bandswitch  (Si)  at  right. 
Counter  dial  at  center  drives  variable  vacuum 
capacitor  Ct.  Tune-operate  switch  Ss  is  at 
tower  _ right,  balanced  on  left  by  pilot  light. 
Amplifier  is  completely  screened  to  reduce 
spurious  energy  radiation.  Screened  area  on 
top  of  shield  allows  circulation  of  air  around 
tube. 


of  bias  voltage  may  be  set  by  adjusting  po- 
tentiometer Ri. 

A simple  automatic  load  control  circuit  is 
incorporated  in  the  input  circuit  of  the  ampli- 
fier. A diode  samples  the  incoming  signal  and 
produces  a d.c.  signal  proportional  to  the  input 
signal.  This  control  voltage  may  be  applied  to 
an  amplifier  tube  in  the  sideband  exciter  in 
such  a way  as  to  limit  the  incoming  signal  to 
a predetermined  value  which  may  be  set  by 
adjustment  of  capacitor  C3  and  potentiometer 
Rz.  This  circuit  is  analogous  to  the  automatic 
volume  control  circuit  in  a receiver.  The  time 
constant  of  the  a.l.c.  circuit  is  determined  by 
the  0.1  gfd.  capacitor  and  33K  resistor  across 
the  output  jack,  Js. 

Amplifier  The  amplifier  is  built  upon  an 

Consfrucfion  aluminum  chassis  measuring 

13"  X 17"  X 4".  The  front 
panel  is  13M"  high,  and  the  above-chassis 


Figure  32 
TOP  VIEW  OF 
KILOWATT 
AMPLIFIER  USING 
4-400A  TETRODE 

Plate  circuit  inductor  is 
placed  parallel  to  panel 
and  driven  by  right-angle 
drive  unit.  Loading  ca- 
pacitor is  at  right,  with 
variable  vacuum  capaci- 
tor mounted  vertically  at 
rear  of  chassis.  4-400A 
tube  and  air  socket  are 
at  rear  right  corner  of 
chassis,  with  bias  trans- 
former in  left  corner. 
Plate  blocking  capaci- 
tors are  placed  between 
aluminum  plates  and 
mounted  atop  vacuum 
capacitor.  Plate  r-f  choke 
is  hidden  between  tube 
and  variable  vacuum  ca- 
pacitor. At  base  of  ca- 
pacitor are  the  B-plus 
VHF  filter  chokes  and 
bypass  capacitors. 

Meiers  are  isolated  by 
aluminum  shield  running 
full  length  of  chassis. 
Meter  leads  pass  down 
conduit  (left  corner,}  to 
under-chassis  area.  Shield 
plate  covers  entire  en- 
closure. 


Figure  33 
UNDER-CIUSSIS 
VIEW  OF 
CATHODE-DRIVEN 
AMPLIFIER 

Bias  and  screen  supply, 
and  input  circuits  are 
placed  in  the  under- 
chassis area.  Filament 
transformer  and  filament 
choke  are  at  left,  with 
tube  socket  next  to  the 
choke.  Screen  terminals 
of  socket  are  grounded 
with  wide  aluminum 
strap  passing  across  ter- 
minals to  socket  shell 
and  chassis.  Grid  termin- 
al is  bypassed  to  center 
of  strap.  Input  capaci- 
tors mount  between  fil- 
ament terminals  and  co- 
axial input  receptacle  on 
rear  of  chassis. 

At  right  are  power  sup- 
ply compartment  (see 
figure  34)  and  voltage 
regulator  tubes.  Vacuum 
capacitor  gear  drive  and 
output  circuit  voltmeter 
are  at  center  of  chassis. 
Blower  motor  is  next  to 
filament  transformer. 
Under-chassis  area  is 
pressurized  so  that  air 
escapes  through  tube 
socket  vent.  After  bot- 
tom plate  is  in  place, 
seam  around  plate  is 
sealed  with  cellophane 
tape  to  provide  air-tight 
enclosure. 


SSB  Linear  Amplifier  633 


shield  is  954”  high.  This  shield  is  made  of  one 
piece  of  aluminum,  bent  into  a U-shape,  hav- 
ing 1/2 -inch  flanges  on  all  sides.  The  open 
side  of  the  "U”  is  bolted  to  the  front  panel  as 
shown  in  the  photographs.  Elastic  stop-nuts 
are  affixed  to  the  shield  every  IV2”  along  the 
flange,  or  it  may  be  drilled  and  tapped  for 
6-32  machine  screws.  A shield  of  this  type 
can  be  homemade,  or  it  may  be  fabricated  in 
any  large  sheet  metal  shop  for  a small  sum. 

The  three  indicating  meters  are  mounted 
on  the  top  front  of  the  panel  and  are  shielded 
from  the  intense  t.f.  field  of  the  amplifier  by 
a half-box  section  of  shielding  bent  to  fit 
around  the  rear  of  the  meter  cases.  The  meter 
shield  measures  21/4"  deep,  4 14"  high,  and  is 
long  enough  to  make  a close  fit  with  the  outer 
shield  at  each  end.  Meter  leads  pass  from  the 
under-chassis  area  through  the  upper  deck  via 
a short  section  of  V2-inch  pipe,  threaded  at 
each  end  and  bolted  between  the  chasms  deck 
and  the  meter  shield. 

The  major  components  are  mounted  with 
an  eye  towards  panel  symmetry  and  short  leads. 
The  vacuum  tube  socket  and  variable  vacuum 
capacitor  G are  placed  to  the  rear  of  the  chas- 
sis. The  capacitor  is  mounted  in  a vertical  po- 
sition, gear  driven  from  the  under-chassis  area 
(figure  33).  Plate  choke  RFGl  is  placed  be- 
tween the  air  socket  and  the  capacitor.  Direct- 


ly in  front  of  the  tube  is  the  panel  driven  1500 
/t/ifd.  loading  capacitor  G,  mounted  on  two 
heavy  aluminum  brackets,  bringing  its  shaft 
in  line  with  that  of  the  variable  inductor,  L,. 

Plate  inductor  Li  is  affixed  to  the  chassis 
parallel  to  the  front  panel,  and  is  driven 
through  a miniature  right  angle  drive,  as  seen 
in  figure  32,  making  the  control  knobs  of  Li 
and  G occupy  symmetrical  positions  on  the 
panel. 

The  variable  vacuum  capacitor  is  centered 
between  the  sides  of  the  chassis  and  is  panel 
driven  by  a counter  dial  and  a right  angle 
drive  unit.  To  preserve  the  panel  appearance, 
the  counter  dial  is  mounted  with  its  shaft 
about  254”  above  the  lower  edge  of  the  panel. 
The  center-line  of  the  right  angle  drive  unit, 
however,  falls  about  1 54  ” above  the  panel 
edge.  Two  1*4”  brass  gears  are  used  to  couple 
the  shafts,  as  seen  in  figure  34.  The  gears  are 
held  in  position  by  an  angle  plate  cut  from  a 
soft  aluminum  sheet. 

The  Tube  Socket  An  Eimac  air  cooled  socket 
is  used  for  the  amplifier 
tube.  The  small  115  volt  a.c.  blower  motor  is 
mounted  near  the  socket,  and  the  entire  under- 
chassis area  is  sealed  air-tight  with  a bottom 
plate.  A Ya"  wide  aluminum  strap  passes 
across  the  center  of  the  socket,  extending  down 


634  H.F.  Power  Amplifiers 


THE  RADIO 


Figure  34 

CLOSE-UP  OF  POWER  SUPPLY 
COMPARTMENT  AND  GEAR 
DRIVE  SYSTEM 

Off-center  shafts  of  capacitor  and  counter 
dial  are  ganged  by  two  gears  held  in  position 
by  triangular  plate  bolted  to  chassis.  Enclo- 
sure below  holds  power  supply  components. 
Silicon  rectifiers  are  mounted  in  fuse  clips 
attached  fo  enclosure  wall.  All  power  leads 
leaving  enclosure  pass  through  small,  bulk- 
head-type ceramic  capacitors.  Primary  leads 
pass  through  "Hypass*‘  capacitors  on  rear 
wall  of  chassis.  Potentiometers  Pi  and  Rg  are 
in  corner  of  chassis. 


to  the  chassis  on  both  sides.  This  strap  is  bolt- 
ed to  the  two  screen  terminals  of  the  socket, 
and  to  the  two  diagonal  assembly  screws  of  the 
socket  frame.  The  ends  of  the  strip  are  fas- 
tened to  the  chassis,  forming  a low  impedance 
return  path  for  the  screen  circuit.  The  .001 
/ifd.,  5 KV  ceramic  grid  bypass  capacitor  is 
mounted  directly  between  the  ground  strap  and 
the  grid  terminal  of  the  tube  socket. 

The  r.f.  input  jack  Ji  is  mounted  to  the  rear 
wall  of  the  chassis  directly  behind  the  tube 
socket  and  two  .001  /xfd.  ceramic  capacitors 
are  connected  between  the  center  terminal  of 
Ji  and  each  filament  terminal  of  the  tube  sock- 
et. Short  leads  connect  the  filament  terminals 
to  the  filament  choke  coil  L.-  mounted  on  the 
side  wall  of  the  chassis,  directly  above  the  tube 
socket.  The  power  supply  components  are  en- 
cased within  an  auxiliary  shield  measuring 
5"  X IVl"  (figure  34).  All  leads  leaving  this 
enclosure  pass  through  .001  ;ufd.  ceramic  feed- 
through capacitors  mounted  on  the  wall  of  the 
enclosure.  The  four  silicon  rectifiers  are  mount- 
ed in  a multiple  fuse  clip  holder  placed  on  the 
side  wall,  and  the  three  regulator  tubes  are 
mounted  on  the  adjacent  wall. 

Amplifier  Wiring  The  power  circuits  of  the 
and  Tuning  amplifier  may  be  wired 

with  unshielded  wire  since 
the  r.f.  field  beneath  the  chassis  is  extremely 
low.  All  d.c.  leads  are  cabled  and  laced  as  may 
be  seen  in  the  under-chassis  photograph.  Leads 
to  the  meter  deck  pass  through  the  area  above 
the  chassis  in  a short  section  of  V^-inch  alum- 


inum conduit,  and  each  meter  is  bypassed  as 
shown  in  the  schematic.  Small  insulated  tip 
jacks  (J«-Jio)  are  mounted  on  the  rear  of  the 
chassis  to  facilitate  voltage  measurements.  In 
addition,  two  insulated,  closed-circuit  jacks 
(Ji,  Js)  allow  the  voltage  regulator  currents  to 
be  read. 

Choke  coils  RFC-2  are  wound  on  a ^-^-inch 
wooden  dowel  rod  which  is  removed,  leaving 
the  coils  self-supporting.  Plate  circuit  wiring 
above  the  deck  is  done  with  silver  plated  cop- 
per strap. 

When  the  amplifier  wiring  is  completed  the 
power  supply  should  be  checked.  The  regula- 
tor tubes  are  inserted  in  the  correct  sockets  and 
the  current  flowing  through  the  regulator  cir- 
cuit is  measured.  The  1500  ohm  series  resistor 
in  the  power  supply  is  adjusted  to  provide  40 
milliamperes  of  screen  regulator  current.  If  the 
correa  current  value  is  obtained  with  less  ':han 
500  ohms  of  series  resistance,  the  1 gfd.  input 
filter  capacitor  should  be  increased  in  value  to 
8 /ifd.  to  boost  the  supply  voltage.  The  regu- 
lator current  divides  between  the  VR-75  and 
the  bias  adjust  potentiometer  Ri,  and  25  mil- 
liamperes of  regulator  current  is  measured  at 
Js  w'hen  the  supply  is  adjusted  for  40  ma. 
at  jack  J^.  After  final  adjustment  has  been 
completed,  the  variable  series  resistor  in  the 
supply  may  be  replaced  with  a fixed  one  if 
desired.  Finally,  bias  control  potentiometer  Ri 
is  set  for  -60  volts  as  measured  between  J» 
(minus)  and  J»  (plus). 

Switch  Ss  should  be  placed  in  the  "tune” 


HANDBOOK 


Kilowatt  Amplifier  635 


Figure  35 

GENERAL  PURPOSE  AMPLIFIER 
OPERATES  IN  CLASS  A,  B,  OR  C 
MODE 

This  kilowatt  amplifier  employs  a pair  of 
4-250A's  in  a pi-network  circuit.  Mode  of  op- 
eration may  be  set  by  selection  of  proper 
screen  and  bias  voltages.  Grid,  plate,  and 
screen  current  meters  are  mounted  on  plastic 
plate  behind  panel  cut-out,  and  tubes  are 
visible  through  shielded  panel  opening.  Across 
bottom  of  panel  (left  to  right)  are  band- 
switch,  grid  tuning,  plate  tuning,  loading,  and 
primary  power  control  circuits.  Plate  tuning 
knob  is  attached  to  small  counter  dial. 


position  before  voltages  are  applied  to  the 
tube.  A suitable  antenna  load  should  be  placed 
on  the  amplifier;  bias,  screen,  and  plate  volt- 
age are  applied.  The  bias  control  should  be  var- 
ied slightly  to  produce  the  correct  resting  plate 
current  as  indicated  in  figure  30.  Excitation 
should  never  be  applied  without  plate  voltage 
on  the  amplifier,  or  without  an  external  an- 
tenna load,  since  the  resultant  high  grid  and 
screen  current  would  damage  the  tube.  In  ad- 
dition, screen  voltage  should  never  be  applied 
without  plate  voltage.  The  combination  screen 
and  bias  supply  of  the  amplifier  is  designed 
to  be  turned  on  with  the  plate  supply  through 
the  "relay  control”  terminal.  All  voltages  may 
be  left  on  the  amplifier  during  "VOX”  oper- 
ation, or  the  supplies  may  be  turned  on  by  the 
"VOX”  circuit.  The  "tune-operate"  switch  S2 
should  not  be  thrown  when  excitation  is  ap- 
plied to  the  tube,  since  the  momentary  break- 
ing of  the  screen  to  ground  path  will  apply 
the  full  excitation  power  to  the  control  grid  of 
the  tube  during  the  moment  the  switch  is  in 
transition  from  one  position  to  the  other.  In 
practice,  this  switching  action  has  been  done 
with  full  excitation  applied  to  the  tube,  but 
it  is  not  recommended. 

The  exciter  should  now  be  connected  and 
a small  amount  of  carrier  is  inserted.  The  plate 
resonant  point  is  found  with  tuning  capacitor 
G and  the  amplifier  loading  adjustments  are 
made,  gradually  increasing  the  excitation  and 
loading  until  the  single  tone  conditions  out- 
lined in  figure  30  are  met.  Speaking  into  the 
microphone  should  produce  approximately  the 
same  peak  conditions  as  obtained  with  catriet 
insertion.  The  linearity  of  the  amplifier  may  be 
checked  with  the  use  of  two  envelope  detectors 
as  described  in  an  earlier  chapter. 


29-9  Kilowatt 

Amplifier  for  Linear  or 
Class  C Operation 

A pair  of  4-250A  or  4-400A  tetrode  tubes 
may  be  employed  in  a pi-coupled  amplifier 
capable  of  running  one  kilowatt  input,  c-w  or 
plate  modulated  phone,  or  two  kilowatts 
p.e.p.  for  sideband  operation.  Correct  choice  of 
bias,  screen,  and  exciting  voltages  will  permit 
the  amplifier  to  function  in  either  class  A,  B, 
or  class  C mode.  The  amplifier  is  designed 
to  operate  at  plate  potentials  up  to  4000  volts, 
and  excitation  requirements  for  class  C oper- 
ation are  less  than  25  watts. 

A bandswitching  type  of  pi-network  is  em- 
ployed in  the  plate  circuit  of  such  an  ampli- 
fier, shown  in  figure  35.  The  pi-network  is 
an  effective  means  of  obtaining  an  impedance 
match  between  a source  of  r.f.  energy  and  a 
low  value  of  load  impedance.  A properly  de- 
signed pi-network  is  capable  of  transformation 
ratios  greater  than  10  to  1,  and  will  provide 
approximately  30  decibels  or  more  attenuation 
to  the  second  harmonic  output  of  the  amplifier 
as  compared  to  the  desired  signal  output.  Since 
the  second  harmonic  level  of  the  amplifier 
tube  may  already  be  down  some  20  db,  the 
actual  second  harmonic  output  of  the  network 
will  be  down  perhaps  50  db  from  the  funda- 
mental power  level  of  the  transmitter.  Atten- 
uation of  the  third  and  higher  order  harmonics 
will  be  even  greater. 


636  H.F.  Power  Amplifiers 


Figure  36 

SCHEMATIC,  GENERAL  PURPOSE  KILOWATT  AMPLIFIER 


C;*— -700  iilifd.  Hammar~ 
iund  HF-JOO 
200  fx/xfd.,  10KV  vat’ 
table  vacuum  capaci- 
tor. Jennings  UCS-200 
Ca-— 7500  /iixfd.,  variable 
capacitor.  Cardwell 
8013 

Ci — Neutralizing  capaci- 
tor, disc.  Millen  7507  7 
Cs — 300  /j-iifd.,  mica, 

1250  volt 

Li-Lio — See  coil  table, 
figure  39 


PC-~-47  ohm,  2 wait 
composition  resistor 
wound  with  6 turns 
it18e. 

RFCi — 2.5  mh.  choke. 
National  R-100 

RFCi — Heavy  duty,  wide- 
band r.f.  choke.  Bar- 
ker & Williamson  type 
800 

PFC,—VHF  choke.  Ohm- 
ite  Z-144 


S, — Two  pole,  6 position 
switch.  Two  Ceniralab 
PA-17  decks,  with  PA- 
SO 1 index  assembly 

Sg—Two  pole,  6 position 
high  voltage  switch. 
Communication  Prod- 
ucts Co.  type  88  two 
gang  switch 

$3 — Four  pole,  three  po- 
sition switch.  Centra- 
lab 


Ti — 5 voit,  20  ampere. 
Stancor  P-6492 

Mi— 0-50  ma.  d.c. 

Triplett 

M: — 0 - ISO  ma.  d.c. 

Triplett 

Ms — 0 - 750  ma.  d.c. 

Triplett 

Gears — 2 required.  Bos- 
ton Gear  #G-465  and 
ftG-466 


The  peak  voltages  encountered  across  the 
input  capacitor  of  the  pi-network  are  the  same 
as  would  be  encountered  across  the  plate  tun- 
ing capacitor  of  a single-ended  tank  used  in 
the  same  circuit  configuration.  The  peak  volt- 
age to  be  expected  across  the  output  capacitor 
of  the  network  will  be  less  than  the  voltage 
across  the  input  capacitor  by  the  square 
root  of  the  ratio  of  impedance  transformation 
of  the  network.  Thus  if  the  network  is  trans- 
forming from  5000  ohms  to  50  ohms,  the  ra- 
tio of  impedance  transformation  is  100  and 
the  square  root  of  the  ratio  is  10,  so  that  the 
voltage  across  the  output  capacitor  is  1/10 
that  across  the  input  capacitor. 

A considerably  greater  value  of  maximum 
capacitance  is  required  of  the  output  capacitor 
than  of  the  input  capacitor  of  a pi-network 
when  transformation  to  a low  impedance  load 
is  desired.  For  3.5  Me.  operation,  maximum 
values  of  output  capacitance  may  run  from 


500  /tgfd.  to  1500  /i(U.fd.,  depending  upon  the 
ratio  of  transformation.  Design  information 
covering  pi-network  circuits  is  given  in  an  earl- 
ier chapter  of  this  Handbook. 

Illustrated  in  figures  35-41  is  an  up-to- 
date  version  of  an  all-band  pi-network  ampli- 
fier, suited  for  sideband  or  class-C  operation. 
The  unit  is  designed  for  TVI-free  operation 
over  this  range. 

Circuit  The  schematic  of  the  general 

Description  purpose  amplifier  is  shown  in 

figure  36.  The  symmetrical 
panel  arrangement  of  the  amplifier  is  shown 
in  the  front  view  (figure  35)  and  the  rear 
view  (figure  37).  A 200  ggfd.  variable  va- 
cuum capacitor  is  employed  in  the  input  side 
of  the  pi-network,  and  a 1500  ggfd.  variable 
air  capacitor  is  used  in  the  low  impedance  out- 
put side.  The  coils  of  the  network  are  switched 
in  and  out  of  the  circuit  by  a two  pole,  five 


Figure  37 

REAR  VIEW  OF  GENERAL  PURPOSE 
AMPLIFIER  WITH  SHIELD 
REMOVED 

The  pi-network  circuit  is  built  from  an  mex* 
pensive  high  voltage  rotary  switch,  and  five 
inductors.  The  switch  is  panel  driven  by  a 
gear  and  shaft  system  shown  in  figure  38. 
Variable  vacuum  capacitor  is  mounted  ver- 
tically between  the  tubes,  directly  in  back  of 
the  plate  r.f.  choke.  Neutralixing  capacitor  is 
at  right,  connected  to  plates  of  tubes  with  a 
wide,  silver  plated  copper  strap.  Meters  are 
enclosed  by  aiuminum  shield  partition  running 
the  width  of  the  enclosure,  with  conduit  car- 
rying meter  loads  to  under-chassis  area  at 
left,  front  of  chassis.  Metal  shells  of  tube 
bases  are  grounded  by  spring  contacts. 


position  high  voltage  ceramic  rotary  switch. 
Each  coil  is  adjusted  for  optimum  circuit  Q, 
resulting  in  no  tank  circuit  compromise  in  ef- 
ficiency at  the  higher  frequencies.  A close-up 
of  the  tank  circuit  is  shown  in  the  introductory 
photograph  at  the  chapter  head,  and  complete 
coil  data  is  given  in  figure  39-  The  plate 
blocking  capacitor  is  made  of  two  .001  g,fd., 
5 KV  ceramic  capacitors  connected  in  series. 

Special  precautions  ate  taken  to  insure  op- 
erating stability  over  the  complete  range  of 
amplifier  operation.  The  screen  terminals  of 
each  tube  socket  ate  jumpered  together  with 
i/s"  copper  strap  and  a parasitic  choke  (PC) 
is  inserted  between  the  center  of  the  strap  and 
the  screen  bypass  capacitor.  In  addition,  sup- 


pressor resistors  are  placed  in  the  screen  leads 
after  the  bypass  capacitor  to  isolate  the  sensi- 
tive screen  circuit  from  the  external  power 
leads.  A third  parasitic  choke  is  placed  between 
the  grid  terminals  of  the  tubes  and  the  tuned 
grid  circuit. 

The  five  coils  of  the  grid  circuit  are  en- 
closed in  a small  aluminum  shield  placed  ad- 
jacent to  the  tube  sockets  (figure  38  and  figure 
40).  The  amplifier  is  neutralized  by  a capaci- 
tive bridge  system  consisting  of  neuttalizing 


Figure  38 
PLACEMENT  OF 
PARTS  IN  UNDER- 
CHASSIS AREA 

Grid  tuned  circuit  is  en- 
closed in  separate  enclo- 
sure at  left.  Bondswifch 
projects  out  the  rear  of 
case,  and  is  gear  driven 
by  same  shaft  that  act- 
uates the  plate  band- 
switch.  Switches  ore  driv- 
en through  right-angle 
gear  drives  and  gears. 
Output  capacitor  of  pi- 
network  is  shielded  from 
rest  of  under-chassis 
components. 

The  screen  terminals  of 
each  tube  socket  are 
strapped  together  with 
ys"  copper  ribbon,  and 
low  inductance  screen 
bypass  capacitor  is 
grounded  i o socket 
mounting  bolt.  Screen 
parasitic  choke  mounts 
between  strap  and  ca- 
pacitor terminal.  All 
power  leads  beneath  the 
chassis  are  ran  in  shield- 
ed braid,  grounded  to 
chassis  at  convenient 
points.  B-plus  lead  is 
made  of  section  of  RG- 
a/U  coaxial  cable,  with 
outer  sheath  and  braid 
removed. 


638  H.F.  Power  Amplifiers 


THE  RADIO 


FIGURE  39 

COIL  TABLE  FOR  K I LOWATT  AMPLI  F[  E R 

GRID  COILS 

Ll  - (80  METERS)  ; 46  TURNS  P‘24  £,  3^4”  DIA-,  1 " LONG  ON 
AMPHENOL  POLYSTYRENE  FORM. 

L2-{40METERS):  30  TURNS  P 24  E,  3^4"  0/ A.,  3^4"  LONG 
ON  AMPHENOL  POLYSTYRENE  FORM. 

L3-(20  meters)  : 12TURNS,  Bt  W 30T 1 M I N I DUCTOR, 

3^4"  O/A.,  3^4"  LONG. 

L4-(15  METERS):  7 TURNS,  BSW  3010  M/N/DUCTOR, 

3/4"  OTA.,  TYS"  LONG. 

LS-f^O  meters):  STURNS,  ASABOYE. 

ALL  COILS  HAVE  3 TURN  LINKS  MADE  OF  HOOKUP  WIRE. 


PLATE  COI  LS 

L6-  (80  METERS)  : 17  TURNSP  10,  J^OD.,  6 TURNS  PER  INCH 

y—’A  IR-DUX- 

L7-  (40  METERS)  : tO  TURNS  # 10,  3"  00.,  S TURNS  PER  INCH 
L8-{Z0  METERS)'  9 TURNS,  3Y16"  COPPER  TUBING, 

2 lYZ"  O.D.,  3"  LONG. 

L9-(15  METERS);  7 TURNS,  1X4"  COPPER  TUBING, 

2 1Y4‘  0.  O.,  3 •'  LONG. 

Ll0-(10  meters);  S turns,  1 /4  - copper  tubing, 

2 1X4  " O.O.,  3"  LONG. 


capacitor  Ci  and  Ca,  the  grid  circuit  bypass  ca- 
pacitor. 

Screen  voltage  may  be  removed  for  tune-up 
purposes  by  control  switch  Sa,  section  B.  The 
screen  circuit  is  grounded  in  the  "off”  and 
"fil”  positions  by  means  of  switch  section  C. 

Amplifier  The  complete  amplifier  is 

Construction  built  upon  an  aluminum  chas- 

sis measuring  13”  x 17"  x 3" 
and  has  a 14"  standard  relay  rack  panel.  The 


Figure  40 

GRID  TANK  CIRCUIT  ASSEMBLY 

Coils  are  mounted  to  the  ceramic  switch  decks 
by  their  leads.  A small  aluminum  plate  at- 
tached to  rear  of  the  switch  assembly  rods 
supports  grid  tuning  capacitor  which  projects 
out  rear  of  shielded  enclosure.  Entire  assembly 
may  be  pre-wired  before  placing  in  enclosure. 


grid  drcxiit  components  are  mounted  within 
an  aluminum  box  measuring  3"  x 4"  x 4". 
Plate  loading  capacitor  Ca,  r.f.  choke  RFC-1, 
and  output  connector  J-  are  placed  within  an 
enclosure  measuring  6"  x 6"  x 3",  made  up 
of  aluminum  angle  sections  and  sheet  material. 
The  plate  circuit  shielding  is  made  of  Reynold’s 
"Do-it-yourself”  aluminum  stock,  available  at 
most  hardware  stores. 

Layout  of  the  major  components  can  be  seen 
in  figure  37.  The  two  tube  sockets  are  placed 
directly  behind  the  panel  opening,  with  the 
plate  r.f.  choke  between  them,  and  the  variable 
vacuum  capacitor  is  mounted  vertically  to  the 
chassis  directly  behind  the  sockets,  on  the  cen- 
ter line  of  the  chassis.  To  the  right  of  the  sock- 
ets is  neutralizing  capacitor  Cj.  The  high  volt- 
age ceramic  coil  switch  S^A-B  is  placed  directly 
behind  the  vacuum  capacitor,  mounted  in  a 
vertical  position. 

The  variable  vacuum  capacitor  is  panel  driv- 
en by  a counter-type  dial,  through  a miniature 
right  angle  gear  drive,  as  seen  in  the  under- 
chassis view  (figure  38).  The  plate  and  grid 
band  switches  are  ganged  and  switched  in  uni- 
son by  means  of  a shaft  acting  through  two 
right  angle  gear  drives  and  two  bevel  gears. 
Both  circuits  are  thus  switched  by  the  "Band- 
switch”  control  located  in  the  lower  left  corner 
of  the  front  panel. 

It  is  necessary  to  apply  forced  air  to  the 
sockets  of  the  amplifier  tubes.  A large  115 
volt  a.c.  operated  blower  is  therefore  mounted 
in  the  center  of  the  bottom  shield  plate.  The 
under-chassis  area  is  thus  pressurized  and  the 
majority  of  the  air  escapes  through  the  socket 
ventilation  holes  located  near  the  pins  of  the 
tubes. 

All  wiring  beneath  the  chassis  (with  the 
exception  of  the  filament  leads ) is  done  with 
5KV  insulated  wire,  encased  in  metallic  braid 
which  is  grounded  to  the  chassis  every  inch  or 
so.  The  B-plus  wiring  from  the  high  voltage 
terminal  to  the  plate  current  meter  is  done 
with  a section  of  RG-8/U  coaxial  line  from 
which  the  outer  braid  has  been  removed.  A 
similar  piece  of  line  is  run  from  the  meter  to 
the  plate  r.f.  choke,  RFC-2. 

The  three  meters  are  mounted  upon  a In- 
cite sheet  placed  behind  a second  Incite  sheet 
mounted  behind  a cut-out  in  the  front  panel. 
The  meters  are  shielded  from  the  plate  circuit 
of  the  amplifier  by  an  aluminum  enclosure 
that  covers  the  wiring  and  meters,  running  the 
full  length  of  the  chassis.  The  meter  leads  pass 
through  the  plate  circuit  area  via  a short  length 
of  lA-inch  aluminum  conduit  that  is  threaded 


HANDBOOK 


Kilowatt  Amplifier  639 


Figure  41 

OPERATING  DATA  AND  SCHEMATIC, 
SCREEN  AND  BIAS  SUPPLY 

Ti — 870-410-0-410-870  volts  at  ISO  ma.  and 
60  ma.  5 volts,  2 a.,  6.3  v.  3.5  o. 
Stancor  P-8307 

Ts — 235-0-235  volts  at  40  ma.  Stancor  PC- 
8401 

CHi — 7 henry  at  ISO  ma.  Stancor  C-I7I0 
CHj — 7 henry  at  SO  ma.  Stancor  C-1707 
RYi — Overload  relay,  adjustable  700-250  mo. 
Note:  J:  is  insulated  from  chassis. 


at  each  end  and  bolted  to  the  chassis  and  the 
meter  shield.  Plate  circuit  wiring  above  the 
chassis  is  done  with  Vi-inch  silver  plated  cop- 
per strap  as  seen  in  the  introductory  photo- 
graph at  the  beginning  of  this  chapter. 

Amplifier  After  the  amplifier  is  wired 

Neutralization  and  checked,  it  should  be 
neutralized.  This  operation 
can  be  accomplished  with  no  power  leads  at- 
tached to  the  unit.  The  tubes  are  placed  in 
their  sockets,  and  about  10  watts  of  30  Me. 
r.f.  energy  is  fed  into  the  p/ate  circuit  of  the 
amplifier,  via  the  coaxial  output  plug  The 
plate  and  grid  circuits  are  resonated  to  the 


frequency  of  the  exciting  voltage  with  the  aid 
of  a grid-dip  meter.  Next,  a sensitive  r.f.  volt- 
meter, such  as  a 0-1  d-c  milliammeter  in 
series  with  a 1N34  crystal  diode  is  connected 
to  the  grid  input  receptacle  (Ji)  of  the  am- 
plifier. The  reading  of  this  meter  will  indicate 
the  degree  of  unbalance  of  the  neutralizing 
circuit.  Start  with  a minimum  of  applied  r.f. 
excitation  to  avoid  damaging  the  meter  or  the 
diode.  Resonate  the  plate  and  grid  circuits  for 
maximum  meter  reading,  then  vary  the  setting 
of  neutralizing  capacitor  C4  until  the  reading 
of  the  meter  is  a minimum.  Each  change  in  O 
should  be  accompanied  by  re-resonating  the 
grid  and  plate  tank  circuits.  When  a point  of 
minimum  indication  is  found,  the  capacitor 
should  be  locked  by  means  of  the  auxiliary  set 
screw. 

Complete  neutralization  is  a function  of  the 
efficiency  of  the  screen  bypass  system,  and 
substitution  of  other  capacitors  for  those  noted 
in  the  parts  list  is  not  recommended.  Mica, 
disc-type,  or  other  form  of  bypass  capacitor 
should  not  be  substituted  for  the  units  speci- 
fied, as  the  latter  units  have  the  lowest  value  of 
internal  inductance  of  the  many  types  tested 
in  this  circuit. 

Bios  and  The  amplifier  requires  -6O  to 

Screen  Supply  -110  volts  of  grid  bias,  and 
plus  300  to  600  volts  of 
screen  potential  for  optimum  characteristics 
when  working  as  a class  ABl  linear  ampli- 
fier. Screen  voltage  for  class  C operation 
(phone)  is  400  volts.  The  voltage  may  be 
raised  to  500  volts  for  c.w.  operation,  if  de- 
sired, although  the  higher  screen  voltage  does 
little  to  enhance  operation.  Approximately  -120 
volts  cut-off  bias  is  required  for  either  phone 
or  c-w  operation.  A suitable  bias  and  screen 
power  supply  for  all  modes  of  operation  is 
shown  in  figure  41,  together  with  an  operating 
chart  for  all  operating  voltages.  The  supply 
furnishes  slightly  higher  than  normal  screen 
voltage  which  is  dropped  to  the  correct  value 
by  an  adjustable  series  resistor,  Ri.  This  series 
resistor  is  adjusted  for  30  milliamperes  of  cur- 
rent as  measured  in  meter  jack  Ji  when  the 
supply  is  disconnected  from  the  amplifier. 
Series  bias  resistor  R2  is  adjusted  for  the  same 
current  in  jack  J2  under  the  same  conditions. 
The  value  of  proteaive  bias  may  now  be  set 
by  adjusting  potentiometer  R3. 

Additional  bias  is  required  for  class  C oper- 
ation which  is  developed  across  series  resistor 
Ri.  Switch  Si  is  open  for  class  C operation  and 
closed  for  sideband  operation. 

It  is  imperative  that  the  screens  of  the  tet- 


4CX-1000A  ALL  BAND  AMPLIFIER 
PROVIDES  TWO  KILOWATT  P.E.P. 
FOR  SIDEBAND,  OR  ONE  KILOWATT 
FOR  PHONE  C.W.  OPERATION 

Grid,  plate,  and  output  meters  are  across  top 
of  panel.  Counter  dial  at  center  drives  var- 
iable vacuum  tuning  capacitor,  with  loading 
control  directly  beneath  it.  Meter  switch  is 
at  left,  with  bandswitch  at  right.  Entire  am- 
plifier is  screened  to  reduce  unwanted  radia- 
tion. 


circuit  whenever  the  screen  current  reaches 
approximately  100  milliamperes. 


rode  amplifier  tubes  be  protected  from  exces- 
sive current  that  could  occur  during  tuning 
adjustments,  or  during  improper  operation  of 
the  amplifier.  The  safest  way  to  accomplish 
this  is  to  include  an  overload  relay  that  will 
open  the  screen  circuit  whenever  the  maximum 
screen  dissipation  point  is  reached.  Two  4- 
250A  tubes  or  4-400A  tubes  have  a total  screen 
dissipation  rating  of  70  warts,  therefore  relay 
RY-1  should  be  adjusted  to  open  the  screen 


29-10  A 2 KilowaH 

P.E.P.  All-band  Amplifier 

Described  in  this  section  is  an  efficient  all- 
band linear  amplifier  suited  for  SSB,  phone, 
and  C.W.  operation  up  to  the  maximum  legal 
power  input  limit.  A 4CX-1000A  ceramic 
power  tetrode  is  employed  in  the  basic  circuit 
shown  earlier  in  this  chapter  in  figure  IIC. 

The  Eimac  4CX-1000A  is  a ceramic  and 


H5V  +SCR-  B-  GND, 


Figure  43 

SCHEMATIC,  4CX-T000A  LINEAR  AMPLIFIER 

C; — 500  u/ifd.,  10KV  variable  vacuum  capacitor.  Jennings  UC5L  4-500 
Cs — 7500  /ijifd.  Cardwell  8013 

Li-Si — High‘C,  all-band  turret.  Barker  & Williamson  852 

RFCi — 2.5  mb.  National  R-100 

RFC2 — Heavy  duty  plate  choke.  Raypar  RL-100 

T I — 6 volts,  12.5  amperes,  with  primary  taps.  Stancor  P-6463 

Tr—125  volts,  15  ma.  Stancor  PS-841S 

R] — 700  ohm,  20  watt  non-inductive  resistor.  May  be  made  of  a number  of  2 or  4 watt  composition 
resistors  connected  in  series  parallel.  Two  Sprague  type  NIT,  10  watt  *'Cool-of7m"  200  ohm  resistors 
in  parallel  used. 

Ml,  Ms — 0 “ 7 d.c.  milliammeter  M: — 0 - 7 d.c.  ammeter  Eimac  SK-800  air  socket 


All-band  Amplifier  641 


Figure  44 

END  VIEW  OF  AMPLIFIER 

4CX-'I000A  sub-chassis  is  shown,  with  ptate 
circuit  choke  mounted  at  rear.  External  blow- 
er provides  forced  air  to  cool  tube  seats  via 
entrance  hole  in  rear  of  chassis.  Bias  adjust- 
ment potentiometer  is  in  foreground,  between 
tube  and  meter  switch. 


metal,  forced  air-cooled,  radial  beam  tetrode 
with  a rated  maximum  plate  dissipation  of 
1000  watts.  It  is  a low  voltage,  high  current 
tube  specifically  designed  for  class  ABl  r-f 
linear  amplifier  service  where  its  high  gain 
and  low  distortion  characteristics  may  be  used 
to  advantage.  At  its  maximum  rated  plate 
voltage  of  3000,  it  is  capable  of  producing 
1680  watts  of  single-tone  output  power  in  SSB 
service.  Maximum  grid  dissipation  of  the  4CX- 
lOOOA  is  zero  watts.  The  design  features  which 
make  the  tube  capable  of  maximum  power  op- 
eration without  driving  the  grid  into  the  posi- 
tive region  also  make  it  necessary  to  avoid  posi- 
tive grid  operation. 

Circuit  The  circuit  of  the  all-band  am- 

Deseription  plifier  is  shown  in  figure  43. 

A resistance  loaded,  untuned 
grid  input  circuit  is  employed  in  conjunction 
with  a pi-network  plate  output  circuit.  Grid 
driving  requirements  are  60  volts  peak  across 
resistor  Ri  which  has  a value  of  100  ohms. 
This  corresponds  to  36  watts  p.e.p.  all  of  which 
is  dissipated  in  resistor  Ri.  For  normal  voice 
operation,  a 20  watt  non-inductive  resistor  may 
be  used.  A simple  1N66  diode  voltmeter  is 
incorporated  in  the  grid  circuit  to  monitor  the 
excitation  voltage.  Grid  and  screen  current  are 
also  measured.  The  d.c.  grid  bias  voltage  is 
obtained  from  a small  selenium-type  half-wave 
power  supply  contained  within  the  amplifier. 

The  plate  circuit  is  designed  around  the  new 
Barker  & Williamson  type  852  bandswitching 
turret,  especially  designed  for  high  current, 
low  voltage  tubes  of  this  type.  The  input  tun- 
ing capacitor  of  the  network  is  a 500  /tgfd. 
variable  vacuum  unit,  and  the  output  loading 
control  is  a 1500  ggfd.  variable  air  capacitor. 
For  operation  into  low  impedance  loads  in  the 
80  meter  band,  an  additional  1500  ggfd.  fixed 
mica  capacitor  may  be  placed  in  parallel  with 
the  output  loading  capacitor,  G.  A second 
1N66  diode  voltmeter  is  placed  across  the  out- 
put circuit  as  a tuning  aid. 


Screen  Grid  Tetrode  tubes  may  exhibit  re- 
Operotion  versed  (negative)  screen  current 
to  a greater  or  lesser  degree  de- 
pending upon  individual  tube  design.  This 
characteristic  is  prominent  in  the  4CX-1000A 
and,  under  some  operating  conditions,  indicat- 
ed negative  screen  current  of  the  order  of  25 
milliamperes  may  be  encountered.  In  general, 
high  values  of  negative  screen  current  wiU  be 
measured  at  the  lower  values  of  plate  potential. 

It  is  difficult  to  measure  screen  dissipation 
in  the  presence  of  secondary  emission  of  this 
type.  The  usual  product  of  current  and  voltage 
may  not  provide  an  accurate  picture  of  actual 
screen  dissipation.  Experience  has  shown  that 
the  screen  will  operate  within  the  limits  es- 
tablished for  this  tube  if  the  indicated  screen 
current,  plate  voltage,  and  drive  voltage  ap- 
proximate the  values  given  in  the  operating 
chart  of  figure  48. 

The  screen  supply  voltage  must  be  main- 
tained constant  for  any  values  of  negative  and 
positive  screen  current  that  may  be  encount- 
ered. Dangerously  high  plate  currents  may  flow 
if  the  screen  power  supply  exhibits  a rising 
voltage  characteristic  with  negative  screen  cur- 


642  H.F.  Power  Amplifiers 


rent.  Stabilization  may  be  aaomplished  in  sev- 
eral ways.  A combination  of  voltage  regulator 
tubes  may  be  connected  across  the  screen  sup- 
ply, or  a heavy  bleeder  resistor  may  be  placed 
across  the  supply  that  will  draw  approximately 
70  milliamperes  at  the  operating  potential  of 
the  screen.  Screen  voltage  is  not  particularly 
critical,  especially  at  plate  potentials  of  2500 
volts  and  higher.  However,  the  screen  volthge 
should  be  adjusted  for  best  linearity  of  the 
stage  when  the  tube  is  operated  at  2000  volts 
plate  potential.  A 350  volt,  150  ma.  supply 
utilizing  choke  input  and  bled  to  70  milliam- 
peres, in  conjunction  with  a 100  watt  primary 
"Variac”  or  variable  voltage  transformer  is  a 
satisfactory  source  of  screen  voltage.  Three  VR- 
105  regulator  tubes  passing  40  ma.  may  be 
employed  for  a fixed-voltage  screen  supply. 


Amplifier  Layout  of  the  amplifier  may 

Construction  be  seen  in  figures  42,  44,  45, 
46,  and  47.  A 121/4"  x 19" 
rack  panel  is  employed  to  form  the  front  of  an 
enclosure  14"  deep.  The  enclosure  is  built  of 
■4-inch  aluminum  angle  stock  and  the  top, 
sides  and  back  are  covered  with  perforated 
aluminum  sheet.  The  bottom  of  the  enclosure 
is  formed  from  a solid  sheet  of  aluminum. 
Only  a very  small  chassis  is  required  to  mount 
the  4CX-1000A  tube  socket,  filament  trans- 
former, and  grid  circuit  components.  A chassis 
measuring  6"  x 3"  x 14"  with  one  end  cut 
off  is  employed  (figure  47).  This  chassis  is 
mounted  at  one  end  of  the  enclosure  and  sup- 
ports a small  aluminum  angle  bracket  holding 
the  two  pi-network  capacitors.  The  band- 
switching plate  tank  inductor  is  mounted  di- 


Figure  45 
TOP  VIEW  OF 
4CX-1000A 
AMPLIFIER 

Tetrode  tube  fits  in  f/- 
mac  air  socket  on  small 
chassis  at  one  end  of 
enclosure.  Pi-network 
capacitors  are  mounted 
on  aluminum  bracket  at- 
tached to  chassis,  and 
are  panel  driven  through 
flexible  shaft  couplers. 
Heavy  duty,  high-C  band- 
switching inductor  is  at 
opposite  side  of  assem- 
bly, with  output  circuit 
diode  voltmeter  mounted 
above  it  on  rear  of  en- 
closure. 

Extremely  low  harmonic 
content  of  amplifier  re- 
moves need  of  meter 
shields.  Amplifier  may  be 
run  without  enclosure 
with  no  sign  of  TV  I on 
any  band  in  normal  TV 
signal  area. 


Figure  46 

REAR  VIEW  OF  CHASSIS  ASSEMBLY 

Air  inlei,  power  receptacle  and  coaxial  input 
jack  are  located  on  rear  of  amplifier  chassis. 
Pi-network  capacitor  assembly  is  at  left,  with 
Raypar  wideband  plate  r-f  choke  in  fore- 
ground. Plate  blocking  capacitor  is  made  of 
four  500  /j-Mfd.  TV-type  capacitors  mounted 
in  "sandwich"  between  stator  plate  of  vacuum 
capacitor  and  round  aluminum  plate. 


rectly  to  the  aluminum  bottom  plate  of  the 
enclosure  as  seen  in  figure  45. 

The  low  impedance  plate  tank  circuit  re- 
quires a higher  than  normal  value  of  plate 
coupling  capacitor.  Four  500  fififd.,  20  KV 
TV-type  capacitors  are  therefore  mounted  be- 
tween two  aluminum  plates  to  form  a 2000 
/i/ifd.  "sandwich”  which  is  bolted  to  tbe  rear 
stator  terminal  of  tuning  capacitor  G.  The  10 
meter  inductor  section  of  coil  Li  bolts  directly 
to  the  aluminum  plate,  as  does  the  plate  lead 
of  the  4CX-1000A. 

Special  breechlock  terminal  surfaces  are  used 
on  the  tube  in  place  of  the  conventionrl  base. 
Three  tabs  protrude  through  the  ceramic  en- 
velope for  each  tube  element,  providing  a low 
inductance  termination.  The  new  Eimac  SK- 
800  socket  having  a built-in  screen  bypass 
capacitor  and  an  air  vent  should  be  employed 
with  this  tube. 

The  rated  heater  voltage  for  the  4CX-1000A 
is  6.0  volts  and  the  operating  voltage,  as  meas- 
ured at  the  socket,  should  be  maintained  with- 
in plus  or  minus  5%  of  this  value  if  long  life 
operation  is  to  be  obtained.  Note  that  the  ca- 
thode and  one  side  of  the  heater  are  internally 
connected. 

Sufficient  cooling  must  be  provided  for  the 
anode  and  ceramic-to-metal  seals  to  maintain 
operating  temperatures  below  200  degrees, 
Centigrade.  At  sea  level,  with  an  inlet  air  tem- 
perature of  20  °C.,  a minimum  of  35  cubic 
feet  per  minute  of  air  flow  is  required.  In  this 
case,  the  chassis  is  pressurized  and  a 50  cubic 
foot/minute  blower  is  connected  to  the  ait  in- 
let on  the  rear  of  the  chassis  by  means  of  a 
short  length  of  automobile  radiator  hose.  The 
cooling  air  must  be  maintained  on  the  ceramic 
seals  during  standby  periods  when  only  heater 
voltage  is  applied  to  the  tube. 


Figure  47 

UNDER-CHASSIS  VIEW  OF 
4CX-1000A  AMPLIFIER 

Bias  supply  is  placed  at  one  end  of  chassis. 
Filament  transformer  is  mounted  next  to  air 
socket  and  short,  heavy  strap  leads  are  used 
to  eliminate  voltage  drop.  Special  air  socket 
has  built-in  screen  bypass  capacitor  to  ensure 
low  inductance  ground  path. 


Amplifier  Amplifier  operating  characteris- 
Operation  tics  are  given  in  figure  48.  The 
bias  potentiometer  R2  is  adjusted 
to  provide  a zero-signal  plate  current  of  250 
milliamperes.  At  this  value,  the  bias  voltage 
will  be  approximately  -55  volts.  Resting  plate 
current  and  screen  voltage  may  both  be  varied 
over  small  limits  to  provide  the  greatest  lin- 
earity range.  Maximum  single  tone  plate  cur- 
rent is  1.0  ampere.  Good  linearity  may  be  ob- 
tained at  2000  watts  p.e.p.  at  a plate  voltage  of 
2500  and  maximum  plate  current  of  800  ma. 

Care  should  be  taken  not  to  apply  excitation 
to  the  amplifier  in  the  presence  of  screen  volt- 
age but  with  no  plate  voltage  present.  A relay 
interlock  may  be  applied  to  the  screen  circuit 
to  open  it  in  the  absence  of  plate  voltage.  Grid 
exciting  voltage  should  be  limited  to  about  60 
volts,  and  grid  current  should  be  monitored  to 
see  that  it  does  not  exceed  0.5  milliamperes 
daring  tune-up. 


644  H.F.  Power  Amplifiers 


THE  RADIO 


FIGURE  46  1 

OPERATINGCHARACTERISTICS,  4CX-1000A  AMPLIFIER] 

PLATE  VOLTAGE 

2000 

2500 

3000 

SCREEN  VOLTAGE 

325 

325 

325 

GRID  VOLTAGE 

-55 

-55 

-55 

ZERO  SIGNAL:  iPfAy/}.) 

250 

250 

250 

SINGLE  TONE  Ip(M/I.) 

1000 

1000 

900 

ZERO  SIGNAL;  l0  2(Myt.) 

-2 

-2 

-2 

SINGLE  TONE'.  IG2  (M/i.) 

30 

30 

30 

SINGLE  TONE  PEAK 

R-F  GR  1 D VOLTS 

55 

55 

55 

SINGLE  TONE  DRIVING 
POWER  (w^rrs) 

0 

0 

0 

SINGLE  TONE  POWER 
OUTPUT  (ivArrs ) 

1080 

1460 

16  80 

NOTE'.  ADJUST  GRID  VOLTAGE  TO  OBTAIN  SPECIFIED 
ZERO-SIGNAL  PLATE  dURRENT. 

29-11  A High  Power 
Push-pull  Tetrode  Amplifier 

The  push-pull  tetrode  amplifier  still  retains 
a high  degree  of  popularity  in  spite  of  the 
trend  towards  single-ended  pi-network  ampli- 
fiers. Regardless  of  the  power  level  of  the 
push-pull  amplifier,  the  physical  design  must 
satisfy  two  important  requirements.  First,  the 
plate  and  grid  circuits  of  the  amplifier  must  be 
symmetrical  with  respect  to  each  tube;  and 
second,  the  input  and  output  circuits  must  be 
isolated  from  each  other.  Finally,  after  the 
physical  layout  of  the  amplifier  has  been  de- 


termined, steps  must  be  taken  to  eliminate  aii 

tendencies  towards  parasitic  oscillation. 

Described  in  this  section  is  a push-pull  am- 
plifier intended  for  operation  in  the  13-15 
Me.  region.  Push-pull  4- 1000 A tubes  are  em- 
ployed, running  at  their  maximum  rated  input. 
Other  tubes,  such  as  the  4-400A,  4-2 50A,  or 
4-125A  may  be  substituted  in  the  design,  scal- 
ing down  the  various  components  to  suitable 
siaes  for  the  reduced  power  level.  This  ampli- 
fier design  is  suitable  for  use  in  the  3 - 30  Me. 
frequency  region.  The  lower  limit  of  operation 
is  restricted  by  the  amount  of  tuning  capacity 
in  the  grid  and  plate  tank  circuits.  Padding 
capacitors  placed  across  the  variable  tuning  ca- 
pacitors may  be  employed  to  reach  the  lower 
frequency  portions  of  the  HF  spectrum.  The 
upper  frequency  limit  of  operation  is  deter- 
mined by  the  residual  screen  lead  inductance 
of  the  particular  tubes  used  in  the  circuit.  At 
frequencies  up  to  30  Me.,  no  neutralizing  dif- 
ficulties are  encountered  with  any  of  the  pre- 
viously mentioned  tubes. 

Circuit  The  complete  schematic  of  the 

Description  push-pull  tetrode  amplifier  is 

shown  in  figure  50.  Grid,  fila- 
ment, and  auxiliary  metering  circuits  of  the 
amplifier  are  enclosed  beneath  the  chassis  in 
an  "electrically  tight”  box.  All  leads  leaving 
this  box  are  suitably  filtered  to  prevent  leakage 
of  energy  into  (or  out  of)  the  enclosure.  The 
plate  circuit  of  the  amplifier  is  contained  in  a 
separate  enclosure  above  the  main  chassis. 

A small  degree  of  feedback  exists  within 
the  tetrode  tubes  due  to  the  minute  value  of 
grid-plate  capacity.  It  is  therefore  necessary  to 
cross-neutralize  the  tubes  by  means  of  two 
capacitor  tabs  mounted  near  the  envelope  of 
the  tube  (figure  53).  These  capacitors  are  ad- 
justed for  minimum  r-f  feedback  at  the  high- 
est operating  frequency  of  the  amplifier. 

Any  tendency  towards  parasitic  oscillation 
is  eliminated  by  the  use  of  small  parasitic 
chokes  placed  in  the  grid  and  screen  leads  of 
each  tetrode  tube.  Excellent  parasitic  suppres- 
sion is  achieved  by  strapping  the  screen  ter- 


Figure  49 

HIGH  POWER  PUSH-PULL  AMPLIFIER 
USES  4-1000A  TUBES 

This  amplifier  design  is  typical  of  push-pull 
stages  regardless  of  power  level . Tubes  are 
visible  through  screened  openings  in  panel. 
Controls  are  (top  to  bottom):  Output  link 
tuning,  variable  link  coupling,  plate  circuit 
tuning,  and  grid  circuit  tuning.  Amplifier  is 
housed  in  special  21"  rack.  Filament  meters 
for  each  tube  are  on  auxiliary  panel  holding 
bias  and  screen  supply. 


HANDBOOK 


P-P  Tetrode  Amplifier  645 


C2  4-1000A/4-400A 


^(1^)  - 
-J}^  ^ ( 


n5v.'\.  -BIAS 


C7( 

r 

^Ca 

a- 

C9 

+SCREEN 


NOTE: 

1.  SCREEN  BYPASS  CAPACITORS 
ARE  CENTRALAB  Typ£  SS8 


© i 


Figure  50 

SCHEMATIC.  PUSH-PULL  HIGH  POWER  AMPLIFIER 

Ci-^100  - 100  lififd.  "butter-Hy'*  capacitor.  Barker  & Williamson  type  JCX 

C2,  Cj-— Neufro/izing  capacitors.  1"  x 3''  aluminum  plate  mounted  near  tube 
envelope  (figure  53) 

C.A-B — 60  - 60  /Aiifd.  split’Stator  variable  vacuum  capacitor.  Bimae  VVC-60-20.  See 
text  for  substitute 

Ci — 500  tJ-^ifd.  Johnson  S00C70 

C^-C5- — 0.7  fifd.,  600  volt  feedthrough  capacitor.  Centralab  "HY-pass'* 

RFC, — VHF-iype  choke.  National  R-60.  or  Ohmite  1-50 

RFC; — Heavy  duty  plate  choke.  Surplus  choke  from  AN/ART- 73  transmitter,  or 
Barker  & Williamson  type  800 

PC — Parasitic  suppressor.  3 turns  if  14  wire,  J/a-/nc/»  diam.  in  parallel  with  47  ohm, 
5 watt  carbon  "Globar"  resistor 


minals  of  the  tetrode  socket  together  (figure 
52)  and  inserting  a parasitic  choke  between 
the  screen  terminal  of  the  socket  and  the  screen 
bypass  capacitor. 

Eimac  air  sockets  should  be  employed  with 
the  larger  tubes,  and  a blast  of  air  directed  at 
the  base  seal  of  each  tube  is  required.  Two 
small  115-volt  blower  motors  mounted  on  the 
rear  of  the  chassis  direct  air  streams  through 
sections  of  radiator  hose  to  the  two  tube  sock- 


Figure  51 

REAR  VIEW  OF  AMPLIFIER 

Plate  tank  circuit  is  placed  between  tetrode 
tubes  which  are  mounted  in  air  sockets.  Split- 
stator  variable  vacuum  capacitor  is  panel 
mounted  and  rotors  are  grounded  to  chassis 
by  heavy  copper  strap.  Tank  coil  is  mounted 
on  10"  ceramic  insulators.  Individual  fila- 
ment transformers  for  the  4-1000A  tubes  are 
at  rear  of  chassis.  Each  tube  requires  7.5  volts 
at  21  amperes. 

Swinging  link  is  panel  driven  through  Millen 
right-angle  drive  unit,  and  flexible,  braided 
leads  connect  link  to  antenna  resonating  ca- 
pacitor and  coaxial  receptacle  mounted  in 
top  screen  of  amplifier. 

Dual  blower  motor  is  mounted  to  rear  of 

chassis.  All  power  leads  (including  filament 
leads  to  meters)  are  brought  out  through 
bulkhead-type  capacitors. 


Figure  53 

CLOSE-UP  OF  PLATE  CHOKE  AND 
NEUTRALIZING  CAPACITOR 

Neutralizing  capacitor  is  made  at  tab  soldered 
to  threaded  rod  mounted  in  insulating  button 
affixed  to  chassis. 


Figure  52 

UNDER-CHASSIS  AREA  OF 
PUSH-PULL  AMPLIFIER 

Blower  motors  are  connected  to  air  sockets 
by  short  lengths  of  radiator  hose.  Filament 
leads  are  run  in  braid,  grounded  at  each  end. 
Shield  plate  is  used  oyer  bottom  of  chassis  to 
proyide  "electrically  tight"  enclosure.  Screen 
terminals  of  each  socket  are  strapped  to- 
gether, and  choke  PC  attaches  to  center  of 
strap. 


ets.  It  is  necessary  to  ground  the  metal  base 
ring  of  each  tube  to  the  chassis,  since  this  ring 
forms  part  of  the  screen  shielding  circuit.  If 
grounding  clips  are  not  furnished  with  the 
tube  socket,  they  may  be  made  from  a short 
length  of  ^-inch  wide  brass  strap  formed  to 
spring  against  the  base  ring. 

Excellent  circuit  symmetry  is  achieved  by 
the  use  of  "butter-fly”-type  mning  capacitors 
in  the  grid  and  plate  circuits.  A 100-watt 
plug-in  coil  may  be  employed  in  the  grid  cir- 
cuit, while  heavy-duty  plug-in  coils  will  suf- 
fice in  the  plate  circuit  up  to  the  kilowatt  level. 

If  extremely  high  plate  voltage  is  to  be 
used,  the  Eimac  dual  vacuum  variable  capacitor 
(VVC-60-20)  should  be  employed  to  prevent 
flash-over.  The  less  expensive  Barker  & Wil- 
liamson "butter-fly”  capacitor  may  be  employed 
for  plate  voltages  of  3500  or  less. 


Amplifier  If  4-250A  or  4-125A  tubes 

Consfrucfion  are  employed,  a 15”  x 17"  x 

6"  chassis  may  be  used.  The 
4- 1000 A tubes  require  a 17"  x 19"  x 6" 
chassis.  Panel  height  is  determined  by  the  com- 
ponents mounted  above  the  chassis  deck.  The 
top  of  the  amplifier  compartment  should  cleat 
the  plate  coil  by  at  least  the  diameter  of  the 
coil  to  prevent  flash-over  and  to  hold  the  coil 
Q to  the  highest  possible  value.  The  amplifier 
enclosure  is  made  of  V2-inch  aluminum  angle 
stock  covered  with  perforated  aluminum.  The 
under-chassis  area  of  the  amplifier  is  sealed 
with  a piece  of  solid  aluminum  sheet. 

Plate  circuit  wiring  is  done  with  1/2 -inch 
silver  plated  copper  strap,  and  for  the  higher 
power  levels  all  conneaions  in  this  circuit 
should  either  be  bolted  or  silver-soldered.  All 
grid  circuit  wiring  is  done  with  #10  copper 
wire,  and  all  low  voltage  d.c.  leads  and  a.c. 
wiring  is  done  in  shielded  braid,  the  braid  be- 
ing grounded  to  the  chassis  at  each  end  of 
the  lead.  A short  length  of  RG-8/U  coaxial 
line  is  used  for  the  connection  between  the  co- 
axial input  plug  and  the  link  circuit  of  the 
grid  coil. 

The  reactance  of  the  antenna  link  is  tuned 
out  by  means  of  a 500  ggfd.  variable  capaci- 
tor placed  in  the  ground  return  of  the  link 
line.  This  capacitor  is  adjusted  for  maximum 
loading  with  the  loosest  possible  degree  of 
link  coupling  to  the  amplifier  tank  coil. 

Amplifier  Excitation  and  grid  bias  are  ap- 

Operafion  plied  to  the  amplifier  and  a 

flashlight  lamp  is  connected  to 
the  link  output  conneaions.  The  neutralizing 
tabs  are  varied  in  position  with  respect  to  the 
envelope  of  the  tube  after  the  grid  and  plate 
circuits  have  been  tuned  to  resonance.  Adjust- 
ments are  made  to  the  capacitors  until  the  de- 
gree of  feedback  is  reduced  to  a minimum. 


CHAPTER  THIRTY 


Speech  and 

Amplitude  Modulation  Equipment 


Amplitude  modulation  of  the  output  of  a 
transmitter  for  radiotelephony  may  be  accom- 
plished either  at  the  plate  circuit  of  the  final 
amplifier,  commonly  called  high-level  AM  or 
simply  plate  modulation  of  the  final  stage,  or 
it  may  be  accomplished  at  a lower  level.  Low- 
level  modulation  is  accompanied  by  a plate- 
circuit  efficiency  in  the  final  stage  of  30  to  45 
per  cent,  while  the  efficiency  obtainable  with 
high-level  AM  is  about  twice  as  great,  running 
from  60  to  80  pet  cent.  Intermediate  values  of 
efficiency  may  be  obtained  by  a combination 
of  low-level  and  high-level  modulation;  cath- 
ode modulation  of  the  final  stage  is  a common 
way  of  obtaining  combined  low-level  and  high- 
level  modulation. 

High-level  AM  is  characterized  by  a require- 
ment for  an  amount  of  audio  power  approxi- 
mately equal  to  one-half  the  d-c  input  to  the 
plate  circuit  of  the  final  stage.  Low-level  mod- 
ulation, as  for  example  grid-bias  modulation  of 
the  final  stage,  requires  only  a few  watts  of 
audio  power  for  a medium  power  transmitter 
and  10  to  15  watts  for  modulation  of  a stage 
with  one  kilowatt  input.  Cathode  modulation 
of  a stage  normally  is  accomplished  with  an 
audio  power  capability  of  about  20  per  cent 
of  the  d-c  input  to  the  final  stage.  A detailed 
discussion  of  the  relative  advantages  of  the 
different  methods  for  accomplishing  amplitude 
modulation  of  the  output  of  a transmitter  is 
given  in  an  earlier  chapter. 


Two  trends  may  be  noted  in  the  design  of 
systems  for  obtaining  high-level  AM  of  the 
final  stage  of  amateur  transmitters.  The  first 
is  toward  the  use  of  tetrodes  in  the  output 
stage  of  the  high-power  audio  amplifier  which 
is  used  as  the  modulator  for  a transmitter.  The 
second  trend  is  toward  the  use  of  a high-level 
splatter  suppressor  in  the  high-voltage  circuit 
between  the  secondary  of  the  modulation  trans- 
former and  the  plate  circuit  of  the  modulated 
stage. 

30-1  Modulation 

Tetrode  In  regard  to  the  use  of  tetrodes, 

Modulators  the  advantages  of  these  tubes 

have  long  been  noted  for  use  in 
modulators  having  from  10  to  100  watts  out- 
put. The  6V6,  6L6,  and  807  tubes  have  served 
well  in  providing  audio  power  outputs  in  this 
range.  Recently  the  higher  power  tetrodes  such 
as  the  4-65A,  813,  4-125A,  and  4-250A  have 
come  into  more  general  use  as  high-level  au- 
dio amplifiers.  The  beam  tetrodes  offer  the 
advantages  of  low  driving  power  (even  down 
to  zero  driving  power  for  many  applications) 
as  compared  to  the  moderate  driving  power  re- 
quirements of  the  usual  triode  tubes  having 
equivalent  power-output  capabilities. 

On  the  other  hand,  beam  tetrode  tubes  re- 
quire both  a screen-voltage  power  supply  and 
a grid-bias  source.  So  it  still  is  expedient  in 
many  cases  to  use  zero-bias  triodes  or  even 


647 


648  Speech  and  A.  M.  Equipment 


THE  RADIO 


low-mu  triodes  such  as  the  304TL  in  many 
modulators  lor  the  medium-power  and  high- 
power  range.  A list  of  suggested  modulator 
combinations  for  a range  of  power  output  capa- 
bilities is  given  in  conjunction  with  several 
of  the  modulators  to  be  described. 

Increasing  the  It  has  long  been  known 

Effective  Modu-  that  the  effective  modu- 
lation Percentage  lation  percentage  of  a 
transmitter  carrying  un- 
altered speech  waves  was  necessarily  limited 
to  a rather  low  value  by  the  frequent  high- 
amplitude  peaks  which  occur  in  a speech 
waveform.  Many  methods  for  increasing  the 
effective  modulation  percentage  in  terms  of 
the  peak  modulation  percentage  have  been 
suggested  in  various  publications  and  subse- 
quently tried  in  the  field  by  the  amateur 
fraternity.  Two  of  the  first  methods  suggested 
were  Automatic  Modulation  Control  and  Vol- 
ume Compression.  Both  these  methods  were 
given  extensive  trials  by  operating  amateurs; 
the  systems  do  give  a degree  of  improvement 
as  evidenced  by  the  fact  that  such  arrangements 
still  are  used  in  many  amateur  stations.  But 
these  systems  fall  far  short  of  the  optimum  be- 
cause there  is  no  essential  modification  of  the 
speech  waveform.  Some  method  of  actually 
modifying  the  speech  waveform  to  improve  the 
ratio  of  peak  amplitude  to  average  amplitude 
must  be  used  before  significant  improvement 
is  obtained. 

It  has  been  proven  thar  the  most  serious  ef- 
fect on  the  radiated  signal  accompanying  over- 
modulation is  the  strong  spurious-sideband  ra- 
diation which  accompanies  negative-peak  clip- 
ping. Modulation  in  excess  of  100  per  cent  in 
the  positive  direction  is  accompanied  by  no 
undesirable  effects  as  far  as  the  radiated  sig- 
nal is  concerned,  at  least  so  long  as  the  linear 
modulation  capability  of  the  final  amplifier  is 
not  exceeded.  So  the  problem  becomes  mainly 
one  of  constructing  a modulator-final  amplifier 
combination  such  that  negative-peak  clipping 
(modulation  in  excess  of  100  per  cent  in  a 
negative  direction)  cannot  normally  take  place 
regardless  of  any  reasonable  speech  input 
level. 

Assymetrical  The  speech  waveform  of  the 
Speech  normal  male  voice  is  charac- 

terized, as  was  stated  before, 
by  high-amplitude  peaks  of  short  duration.  But 
it  is  also  a significant  characteristic  of  this 
wave  that  these  high-amplitude  peaks  are  poled 
in  one  direction  with  respect  to  the  average  am- 
plitude of  the  wave.  This  is  the  "lopsided”  or 


assymetrical  speech  which  has  been  discussed 
and  illustrated  in  an  earlier  chapter. 

The  simplest  method  of  attaining  a high 
average  level  of  modulation  without  negative 
peak  clipping  may  be  had  merely  by  insuring 
that  these  high-amplitude  peaks  always  are 
poled  in  a positive  direction  at  the  secondary 
of  the  modulation  transformer.  This  adjust- 
ment may  be  achieved  in  the  following  man- 
ner: Couple  a cathode-ray  oscilloscope  to  the 
output  of  the  transmitter  in  such  a manner  that 
the  carrier  and  its  modulation  envelope  may 
be  viewed  on  the  scope.  Speak  into  the  micro- 
phone and  note  whether  the  sharp  peaks  of 
modulation  are  poled  upward  or  whether  these 
peaks  tend  to  cut  the  baseline  with  the  "bright 
spot”  in  the  center  of  the  trace  which  denotes 
negative-peak  clipping.  If  it  is  not  obvious 
whether  or  not  the  existing  polarity  is  correct, 
reverse  the  polarity  of  the  modulating  signal 
and  again  look  at  the  envelope.  Since  a push- 
pull  modulator  almost  invariably  is  used,  the 
easiest  way  of  reversing  signal  polarity  is  to 
reverse  either  the  leads  which  go  to  the  grids 
or  the  leads  to  the  plates  of  the  modulator 
tubes. 

When  the  correct  adjustment  of  signal  po- 
larity is  obtained  through  the  above  procedure, 
it  is  necessarily  correct  only  for  the  specific 
microphone  which  was  used  while  making  the 
tests.  The  substitution  of  another  microphone 
may  make  it  necessary  that  the  polarity  be  re- 
versed, since  the  new  microphone  may  be  con- 
nected internally  in  the  opposite  polarity  to 
that  of  the  original  one. 

Low-Level  The  low-level  speech  clip- 

Speech  Clipping  per  is,  in  the  ideal  case,  a 

very  neat  method  for  ob- 
taining an  improved  ratio  of  average-to-peak 
amplitude.  Such  systems,  used  in  conjunction 
with  a voice-frequency  filter,  can  give  a very 
worthwhile  improvement  in  the  effective  mod- 
ulation percentage.  But  in  the  normal  amateur 
transmitter  their  operation  is  often  less  than 
ideal.  The  excessive  phase  shift  between  the 
low-level  clipper  and  the  plate  circuit  of  the 
final  amplifier  in  the  normal  transmitter  re- 
sults in  a severe  alteration  in  the  square-wave 
output  of  the  clipper-filter  which  results  from 
a high  degree  of  clipping.  The  square-wave 
output  of  the  clipper  ends  up  essentially  as  a 
double  saw-tooth  wave  by  the  time  this  wave 
reaches  the  plate  of  the  modulated  amplifier. 
The  net  result  of  the  rather  complex  action  of 
the  clipper,  filter,  and  the  phase  shift  in  the 
succeeding  stages  is  that  the  low-level  speech 
clipper  system  does  provide  an  improvement 


HANDBOOK 


Design  649 


MOD,  R.F,  115  V. 
FINAL 


Figure  1 

HIGH-LEVEL  SPLATTER  SUPPRESSOR 

The  high-vaeuum  diode  acts  as  a series  fim- 
iter  to  suppress  negative-peak  clipping  in 
the  modulated  r-f  amplifier  as  a result  of 
large  amplitude  negative-peak  modulating 
signals.  In  addition,  the  low-pass  filter  fol- 
lowing the  diode  suppresses  the  transients 
which  result  from  the  peak-clipping  action  of 
the  diode.  Further,  the  filter  attenuates  all 
harmonics  generated  within  the  modulator 
system  whose  frequency  lies  above  the  cut- 
off frequency  of  the  filter.  The  use  of  an 
appropriate  value  of  capacitor,  determined 
experimentally  as  discussed  in  Chapter 
Fifteen,  across  the  primary  of  the  modulation 
transformer  (Cs)  introduces  further  attenua- 
tion to  high-frequency  modulator  harmonics. 
Chokes  suitable  for  use  at  L are  manufactured 
by  Chicago  Transformer  Corp..  The  correct 
values  of  capacitance  for  Ci,  C:,  Cs,  and  C < 
are  specified  on  the  installation  sheet  for  the 
splatter  suppressor  chokes  for  a wide  variety 
of  operating  conditions. 

in  the  effective  modulation  percentage,  but  it 
does  not  insure  against  overmodulation.  An 
extensive  discussion  of  these  factors,  along 
with  representative  waveforms,  is  given  in 
Chapter  Fifteen.  Circuits  for  some  recommend- 
ed clipper-filter  systems  will  also  be  found 
in  the  same  chapter. 


High  -Level  One  practicable  method 

Splatter  Suppressor  for  the  substantial  elim- 

ination of  negative-peak 
clipping  in  a high-level  AM  transmitter  is  the 
so-called  high-level  splatter  suppressor.  As 
figure  1 shows  it  is  only  necessary  to  add  a 
high-vacuum  rectifier  tube  socket,  a filament 
transformer  and  a simple  low-pass  filter  to  an 
existing  modulator-final  amplifier  combination 
to  provide  high-level  suppression. 

The  tube,  Vi,  serves  to  act  as  a switch  to 
cut  off  the  circuit  from  the  high-voltage  power 
supply  to  the  plate  circuit  of  the  final  ampli- 
f'er  as  soon  as  the  peak  a-c  voltage  across 
the  secondary  of  the  modulation  transformer 
has  be-ome  equal  and  opposite  to  the  d-c  volt- 
age being  applied  to  the  plate  of  the  final  am- 
plifier stage.  A single-section  low-pass  filter 
serves  to  filter  out  the  high-frequency  compo- 
nents resulting  from  the  clipping  action. 

Tube  Vi  may  be  a receiver  rectifier  with  a 
5-volt  filament  for  any  but  the  highest  power 
transmitters.  The  5Y3-GT  is  good  for  123  ma. 
plate  current  to  the  final  stage,  the  5R4-GY 
and  the  5U4-G  are  satisfactory  for  up  to  250 
ma.  For  high-power  high-voltage  transmitters 
the  best  tube  is  the  high-vacuum  transmitting 
tube  type  836.  This  tube  is  equivalent  in 
shape,  filament  requirements,  and  average-cur- 
rent capabilities  to  the  866A.  However,  it  is 
a vacuum  rectifier  and  utilizes  a large-size 
heater-type  dual  cathode  requiring  a warm-up 
time  of  at  least  40  seconds  before  current 
should  be  passed.  The  tube  is  rated  at  an 
average  current  of  250  ma.  For  greater  current 
drain  by  the  final  amplifier,  two  or  more  836 
tubes  may  be  placed  in  parallel. 

The  filament  transformer  for  the  cathode  of 
the  splatter-suppressor  tube  must  be  insulated 


Figure  3 
UNDERCHASSIS 
OF  THE 

6L6  MODULATOR 

A 4-connector  plug  is 
used  for  filaments  and 
plate  voltage  to  the 
speech  amplifier,  while  a 
6-wire  terminal  strip  is 
used  for  the  high-voltage 
connections  and  the  trans- 
mitter'control  switch. 


for  somewhat  more  than  twice  the  operating 
d-c  voltage  on  the  plate  modulated  stage,  to 
allow  for  a factor  of  safety  on  modulation 
peaks.  A filament  transformer  of  the  type  nor- 
mally used  with  high-voltage  rectifier  tubes 
will  be  suitable  for  such  an  application. 

30-2  Design  of  Speech 
Amplifiers  and  Modulators 

A number  of  representative  designs  for 
speech  amplifiers  and  modulators  is  given  in 
this  chapter.  Still  other  designs  are  included 
in  the  descriptions  of  other  items  of  equipment 
in  other  chapters.  However,  those  persons  who 
wish  to  design  a speech  amplifier  or  modulator 
to  meet  their  particular  needs  are  referred  to 
Chapter  Six,  Vacuum  Tube  Amplifiers,  for  a 
detailed  discussion  of  the  factors  involved  in 


the  design  of  such  amplifiers,  and  for  tabular 
material  on  recommended  operating  conditions 
for  voltage  and  power  amplifiers. 

10  to  120  Wott  It  is  difficult  to  surpass 

Modulator  with  the  capabilities  of  the 

Beam  Power  Tubes  reliable  beam  power 
tube  when  an  audio 
power  output  of  10  to  120  watts  is  required 
of  a modulator.  A pair  of  6L6  tubes  operating 
in  such  a modulator  will  deliver  good  plate 
circuit  efficiency,  require  only  a very  small 
amount  of  driving  power,  and  they  impose  no 
serious  grid-bias  problems. 

Circuit  Included  on  the  chassis  of  the 

Description  modulator  shown  in  figures  2 
and  3 are  the  speech  amplifier, 
the  driver  and  modulation  transformers  for  the 


6SJ7  6J5  6SN7-GT  Ti  Vi  Ta 


Figure  4 

SCHEMATIC  OF  BEAM  POWER  TUBE  MODULATOR 

M — 0 - 250  dx.  milliammeier  Tg — **Poly-pedance"  Modulation  transformer 

60-wati  level  = Stancor  A~3893,  or  UTC  S-20 

Ti — Driver  transformer.  Staneor  A-4701,  or  UTC  S-10.  12S-wait  level  = Stancor  A-3894 


General  Purpose  Modulator  651 


output  tubes,  and  a plate  current  milliammeter. 
The  power  supply  has  not  been  included.  The 
6SJ7  pentode  first  stage  is  coupled  through 
the  volume  control  to  the  grid  of  a 6J5  phase 
inverter.  The  output  of  the  phase  inverter  is 
capacitively  coupled  to  the  grids  of  a 6SN7-GT 
which  acts  as  a push-pull  driver  for  the  output 
tubes.  Transformer  coupling  is  used  between 
the  driver  stage  and  the  grids  of  the  output 
tubes  so  that  the  output  stage  may  be  operated 
either  as  a class  ABi  or  class  AB2  amplifier. 

The  Output  Either  6L6,  6L6-G  or  807 

Stage  tubes  may  be  used  in  the  out- 

put stage  of  the  modulator.  As 
a matter  of  fact,  either  6V6-GT  or  6F6-G 
tubes  could  be  used  in  the  output  stage  if 
somewhat  less  power  output  is  required.  The 
807  tube  is  the  transmitting-tube  counterpart 
of  the  6L6  and  carries  the  same  ratings  and 
recommended  operating  conditions  as  the  6L6 
within  the  ratings  of  the  6L6.  But  the  807 
does  have  somewhat  greater  maximum  ratings 
when  the  tube  is  to  be  used  for  ICAS  (Inter- 
mittent Commercial  and  Amateur  Service)  op- 
eration. The  6L6  and  807  retail  to  the  amateur 
for  essentially  the  same  price,  although  the  807 
is  available  only  from  transmitting  tube  dis- 
tributors. The  6L6-G  tube  retails  for  a some- 
what lower  price;  hence  it  is  expedient  to  pur- 
chase 6L6-G  tubes  if  360  to  400  volts  is  the 
maximum  to  be  used  on  the  output  stage,  or 
807  tubes  if  up  to  750  volts  will  be  applied. 

Tabulated  in  figure  5 are  a group  of  recom- 
mended operating  conditions  for  different  tube 
types  in  the  output  stage  of  the  modulator.  In 
certain  sets  of  operating  conditions  the  tubes 
will  be  operated  class  ABi,  that  is  with  in- 


F IGURE  5 


RECOMMENDED  OPERATING  CONDITIONS  FOR 
MODULATOR  OF  FIG.4  FOR  DIFFERENT  TUBE  TYPES 


TUBES 

VI.V2 

CLASS 

PLATE 

VOLTS 

(C) 

SCREEN 

VOLTS 

(C) 

GRID 

BIAS 

(D) 

PLATE-ro 
PLATE  LOAD 
(OHMS) 

PLATE 

CURRENT 

(MA.) 

POWER 
OUTPUT 
(WATTS ) 

6V6GT 

ABI 

250 

250 

-15 

10,000 

70-80 

10 

6V6GT 

ABI 

265 

265 

-19 

8,000 

70-95 

15 

6L6 

ABI 

360 

270 

-23 

6,600 

1 

85-135 

27 

6L6 

AB2 

360 

270 

-23 

3,800 

85-205 

47 

807 

AB  1 

600 

300 

-34 

10,000 

35-140 

56 

807 

ABi 

750 

300 

-35 

12,000 

30-140 

75 

807 

AB2 



750 

300 

-35 

7,300 

30-240 

120 

creased  plate  current  with  signal  but  with  no 
grid  current.  Other  operating  conditions  speci- 
fy class  ABi  operation,  in  which  the  plate  cur- 
rent increases  with  signal  and  grid  current 
flows  on  signal  peaks.  Either  type  of  operation 
is  satisfactory  for  communication  work. 

30-3  General  Purpose 
Triode  Class  B Modulator 

High  level  class  B modulators  with  power 
output  in  the  125  to  500  watt  level  usually 
make  use  of  triodes  such  as  the  809,  811, 
8005,  805,  or  810  tubes  with  operating  plate 
voltages  between  750  and  2000.  Figures  6, 
7,  and  8 illustrate  a general  purpose  modula- 
tor unit  designed  for  operation  in  this  power 
range.  Figure  8 gives  a group  of  suggested 
operating  conditions  for  various  tubes.  The 
size  of  the  modulation  transformer  will  of 
course  be  dependent  upon  the  amount  of  audio 
power  developed  by  the  modulator.  In  the  case 


Figure  6 

REAR  VIEW  OF  THE 
GENERAL-PURPOSE  MODULATOR 


652  Speech  and  A.  M.  Equipment 


6SJ7 


6N7 


6B4-G  Ti 


Tz 


SCHEMATIC  OF  GENERAL  PURPOSE  MODULATOR 


M — 0 - 500  ma. 

Ti — Driver  transformer.  Stancor  A-4761 
Tg — "Poly-pedance"  Modulation  transformer. 

300  watt  rating,  Stancor  A-3898. 

500  watt  rating,  Stancor  A-3899. 

Ts — 360  - 0 - 360  volts,  150  ma.  Stancor  PC-8410 


Ti-^Suitable  for  tubes  used. 

For  Sll-A's  — 6.3  volt,  8 amp.  Stancor  P'6308 
For  8 10's  = 10  volt,  10  amp.  Stancor  P-6461 
CHi — 14  henry,  700  ma.  UTC  $-79.  Stancor  C-7007 
Ri — 1 Kj  10  watts,  adjustable.  Set  for  plate  current 
of  80  ma,  (no  signal)  to  6B4-G  tubes 
(approximately  875  ohms). 

V],Vg — See  figure  8. 


of  the  500  watt  modulator  (figures  9 and  10) 
the  size  and  weight  of  the  components  require 
that  the  speech  amplifier  be  mounted  on  a 
separate  chassis.  For  power  levels  of  300  watts 
or  less  it  is  possible  to  mount  the  complete 
speech  system  on  one  chassis. 


FIGURE  8 

SUGGESTED  OPERATING  CONDITIONS 
FOR  GENERAL  PURPOSE  MODULATOR 


TUBES 
V1,  V 2. 

^ J 

PLATE 

VOLTAGE 

GR  1 0 
BIAS 
(VOLTS) 

PLATE 

CURRENT 

(MA.) 

PLATE  TO 
PLATE  LOAD 
(OHMS) 

SINE  WAVE 
POWER  OUTPUT 
(WATTS  ) 

700 

0 

70-250 

6,2  00 

120 

81  1- A 

750 
1 

0 

30-350 

5.100 

175 

81  1-A 

1 

1000 

0 

45-350 

7,400 

245 

81  1-A 

1 250 

° 

50-350 

9.200 

310 

81  1-A 

1500 

-4.5 

32-315 

12.400 

340 

805 

1 250 

0 

146-400 

6.7  00 

300 

I 805 

1 500 

-16 

84-400 

8,200 

370 

I 810 

2000 

-50 

60-420 

1 2p00 

450 

810 

2500 

-75 

50-420 

17,500 

500 

* 8005 

1 

1 500 

-67 

40-330 

9,800 



330 

Circuit  Description  The  modulator  unit 

of  General  Purpose  shown  in  figure  6 is 
Modulator  complete  except  for  the 

high  voltage  supply  re- 
quired by  the  modulator  tubes.  A speech  am- 
plifier suitable  for  operation  with  a crystal 
microphone  is  included  on  the  chassis  along 
with  its  own  power  supply.  A 6SJ7  is  used 
as  a high  gain  preamplifier  stage  resistance 
coupled  to  a 6N7  phase  inverter.  The  audio 
level  is  controlled  by  a potentiometer  in  the 
input  grid  circuit  of  the  6N7  stage.  Push-pull 
6B4-G  low  fi  triodes  serve  as  the  class  B driver 
stage.  The  6B4’s  are  coupled  to  the  grids  of 
the  modulator  tubes  through  a conventional 
multi-purpose  driver  transformer.  Cathode  bias 
is  employed  on  the  driver  stage  which  is  capa- 
ble of  providing  12  watts  of  audio  power  for 
the  grid  circuit  of  the  modulator. 

The  modulator  illustrated  in  figures  9 and 
10  is  designed  for  use  with  class  B 810  triode 
tubes  operating  at  a plate  potential  of  2500 
volts.  Maximum  audio  power  available  is  500 
watts,  sufficient  to  100%  modulate  a transmit- 
ter running  one  kw.  input.  The  modulator  may 
be  driven  by  the  simple  speech  amplifier  of  fig- 
ure 12,  or  by  the  clipper-amplifier  of  figure  15. 


Figure  9 

REAR  VIEW  OF  810 
5 00- WATT  CLASS  B 
MODULATOR 
Bias  control  is  directly  be- 
low the  810  tubes  with  bias 
supply  to  right. 


The  modulator  chassis  includes  a small  low 
impedance  bias  supply  capable  of  delivering 
approximately  — 75  volts  under  the  changing 
load  conditions  imposed  by  the  varying  grid 
circuit  impedance  of  the  class  B modulator. 

A low  pass  audio  filter  network  is  placed 
after  the  modulator  stage  to  reduce  harmonics 


generated  in  the  audio  system  which  would 
normally  show  up  as  side-band  "splatter”  on 
the  phone  signal.  Frequencies  above  3500 
cycles  are  attenuated  15  decibels  or  more  by 
the  simple  pi-network  filter  shown  in  figure 
10. 


Figure  1 0 

SCHEMATIC  OF  500  WATT  TRIODE  MODULATOR  WITH  BIAS  SUPPLY 


Tj — Driver  transformer,  15  watts.  Stancor  A~4761 

Ti — Modu/ation  transformer,  500  watt.  Chicago  Transformer  Co.  CMS'3 

Tj — 5 volts,  3 amp.  Stancor  P-6467 

Tj. — Tapped  bias  transformer,  15-100  volts.  UTC  5-51 

Ts — TO  volt,  10  amp.  Stancor  P-6461 

CHi — 500  ma.  "splatter  choke"  Chicago  Transformer  Co.  SR-SOO 
CHs — 7 henry,  ISO  ma.  Stancor  C-1710 
M — 0 - 500  d.c.  milliammeter 
SiA-B — High  voltage  switch  (see  text) 


654  Speech  and  A.  M.  Equipment 


THE  RADIO 


0-  B-t-  TO 

2500  V OSCILLOSCOPE 


Figure  1 1 

TEST  SETUP  FOR  500-WATT 
MODULATOR 


For  c-w  operation  the  secondary  of  the  class 
B modulation  transformer  is  shorted  out  and 
the  filament  and  bias  circuits  of  the  modulator 
are  disabled.  Switch  SiA  must  have  10,000 
volt  insulation  rating.  A suitable  switch  may 
be  found  in  the  war  surplus  BC-306A  antenna 
loading  unit. 

All  low  voltage  connections  to  the  modula- 
tor are  brought  out  to  a six  terminal  phenolic 
strip  on  the  back  of  the  chassis.  A 0-500  ma. 
d-c  meter  is  placed  in  the  filament  circuit  of 
the  810  tubes.  The  meter  is  placed  across  a 
50  ohm,  1 watt  resistor  so  that  the  filament 
return  circuit  of  the  modulator  is  not  broken 
if  the  meter  is  removed. 

The  modulation  transformer,  Ts,  is  designed 
for  plate-to-plate  loads  of  either  12,000  ohms 
or  18.000  ohms  when  a 6250  ohm  load  is 
placed  across  the  secondary  terminals.  The  810 
tubes  are  correctly  matched  when  the  18,000 
ohm  taps  are  used  at  a plate  potential  of 
2500  volts,  or  when  the  12,000  ohm  taps  are 
used  at  a plate  potential  of  2000  volts  or  2250 
volts. 

Modulator  Because  of  the  great  weight  of 
Construction  the  modulator  components  it  is 
best  to  use  a heavy-duty  steel 
chassis.  A 13”  X 17"  x 4”  chassis  (Bud  CB- 
643),  a 14"  steel  panel  (Bud  PS-1257G)  and 
a pair  of  Bud  MB-449  mounting  brackets  make 
up  the  assembly  for  this  particular  modulator. 
As  seen  in  figure  9,  the  CMS-3  modulation 
transformer  is  mounted  in  the  left-front  corner 
of  the  unit.  The  secondary  terminals  of  Tz  are 
to  the  front  of  the  chassis,  clearing  the  front 
panel  by  about  Vi" ■ To  the  right  of  Tz  is 
placed  the  high  level  audio  filter  choke,  CHi. 
The  two  810  tubes  are  mounted  in  back  of  Ti. 
To  the  right  of  the  810  tubes  is  the  10  volt 
filament  transformer,  T-,.  To  the  right  of  Ts  is 
the  5Y3-GT  bias  rectifier  tube.  Between  Ts 
and  CHi  are  mounted  the  mica  bypass  capaci- 


tors which  make  up  the  high  level  filter  net- 
work. Two  .003  gfd.,  5000  volt  mica  capacitors 
are  paralleled  for  Ci  and  also  for  G.  G is  made 
from  a .001  gfd.  capacitor  and  a .0015  gfd.  ca- 
pacitor which  are  connected  in  parallel.  All 
of  these  capacitors  are  mounted  upon  a ply- 
wood bracket  which  insulates  them  from  the 
metal  chassis.  This  prevents  insulation  break- 
down within  the  capacitors  which  might  occur 
if  they  were  fastened  directly  to  the  metal 
chassis. 

The  bias  supply  components  are  mounted 
beneath  the  four-inch  deep  chassis.  Placement 
of  these  parts  is  not  critical.  The  bias  adjust- 
ment control,  Ri,  is  mounted  on  the  back  lip 
of  the  chassis  as  are  the  two  high  voltage  ter- 
minals. Millen  37001  high  voltage  connectors 
are  used  for  the  two  high  voltage  leads.  High- 
voltage  TV  wire  should  be  employed  for  all 
leads  in  the  810  plate  circuit. 

Modulator  When  the  modulator  has  been 
Adjustment  wired  and  checked,  it  should  be 
tested  before  being  used  with 
an  r-f  unit.  A satisfactory  test  set-up  is  shown 
in  figure  11.  A common  ground  lead  should 
be  run  between  the  speech  amplifier  and  the 
modulator.  Six  1000  ohm  100  watt  resistors 
should  be  connected  in  series  and  placed  across 
the  high  voltage  terminals  of  the  modulator 
unit  to  act  as  an  audio  load.  The  first  step  is  to 
place  the  810  tubes  in  their  sockets  and  turn 
switch  Si  to  the  "phone”  position.  The  810 
filaments  should  light,  and  switch  section  Si.t 
should  remove  the  short  across  the  secondary 
of  Tz.  Ri  should  be  adjusted  to  show  — 75 
volts  from  each  810  grid  terminal  to  ground 
as  measured  with  a high  resistance  voltmeter. 
If  an  oscilloscope  is  available,  it  should  be  cou- 
pled to  point  "A”  on  the  load  resistor  (figure 
II)  through  a 500  ggfd.  ceramic  TV  capacitor 
of  10,000  volts  rating.  The  case  of  the  oscil- 
loscope should  be  grounded  to  the  common 
ground  point  of  the  modulator. 

A plate  potential  of  2500  volts  is  now  ap- 
plied to  the  modulator,  and  Ri  adjusted  for 
a resting  plate  current  of  50  milliamperes  as 
read  on  the  500  milliampere  meter  in  the  cath- 
ode circuit  of  the  modulator.  Be  extremely 
careful  during  these  adjustments,  since  the 
plate  supply  of  the  modulator  is  a lethal  weap- 
on. Never  touch  the  modulator  when  the  plate 
voltage  supply  is  on!  Be  sure  you  employ  the 
TV  blocking  capacitor  between  the  oscillo- 
scope and  the  plate  load  resistors,  as  these 
load  resistors  are  at  high  voltage  potential! 
If  a high  resistance  a-c  voltmeter  is  available 
that  has  a 2000  volt  scale,  it  should  be  clipped 


HANDBOOK 


10-Watt  Amplifier-Driver  655 


6B4G 


between  the  high  voltage  terminals  of  the  mod- 
ulator, directly  across  the  dummy  load.  Do  not 
touch  the  meter  when  the  high  voltage  supply 
is  in  operation!  An  audio  oscillator  should  be 
connected  to  the  audio  input  circuit  of  the  ex- 
citer-transmitter and  the  audio  excitation  to 
the  high  level  modulator  should  be  increased 
until  the  a-c  voltmeter  across  the  dummy  load 
resistor  indicates  an  R-M-S  reading  that  is 
equal  to  0.7  (70%)  of  the  plate  voltage  applied 
to  the  modulator.  If  the  modulator  plate  volt- 
age is  2500  volts,  the  a-c  meter  should  indi- 
cate 1750  volts  developed  across  the  6000 
ohm  dummy  load  resistor.  This  is  equivalent 
to  an  audio  output  of  500  watts.  With  sine 
wave  modulation  at  1000  c.p.s.  and  no  speech 
clipping  ahead  of  the  modulator,  this  voltage 
should  be  developed  at  a cathode  meter  cur- 
rent of  about  350  ma.  when  the  plate-to-plate 
modulator  impedance  of  the  modulator  is  18,- 
000  ohms.  Under  these  conditions,  the  oscillo- 
scope may  be  used  to  observe  the  audio  wave- 
form of  the  modulator  when  coupled  to  point 
"A”  through  the  10,000  volt  coupling  capac- 
itor. 

When  the  frequency  of  the  audio  oscillator 
is  advanced  above  3500  cycles  the  output  level 
of  the  modulator  as  measured  on  the  a-c  volt- 
meter should  drop  sharply  indicating  that  the 
low  pass  audio  network  is  functioning  properly. 

With  speech  waveforms  and  no  clipping  the 
modulator  meter  will  swing  to  approximately 
150-  200  milliamperes  under  100%  modula- 


tion at  a plate  potential  of  2500  volts.  With 
speech  waveforms  and  moderate  clipping  the 
modulator  meter  will  swing  to  about  300  ma. 
under  100%  modulation. 

30-4  A 10- Watt 

Amplifier  - Driver 

A simple  speech  amplifier-driver  for  a medi- 
um powered  class  B modulator  is  shown  in  fig- 
ure 12.  The  amplifier  is  designed  to  work  with 
a crystal  microphone.  The  first  stage  utilizes 
a 6SJ7.  The  gain  control  is  between  the  6SJ7 
plate  circuit  and  the  grid  of  the  6J5  second 
stage  amplifier.  The  output  tubes  are  a pair 
of  6B4-G  low-mu  tubes  operating  with  a self- 
bias resistor  in  their  common  filament  return 
circuit.  Operating  in  this  manner  the  6B4’s 
have  an  undistorted  output  of  approximately 
10  watts.  This  is  sufficient  power  to  drive 
most  class-B  modulators  whose  output  is  500 
watts  or  less.  The  driver  transformer  for  cou- 
pling the  plates  of  the  6B4-G  tubes  to  the 
grids  of  the  Class  B stage  is  not  shown,  as 
it  had  been  found  more  convenient  to  locate 
this  transformer  at  the  grids  of  the  modulator 
tubes  rather  than  in  the  speech  amplifier.  The 
correct  transformer  step-down  ratio  for  driving 
most  class  B tubes  has  been  set  down  in  tabular 
form  by  the  various  transformer  manufacturers. 
When  the  driver  transformer  is  purchased  one 
should  be  obtained  which  has  the  proper  turns 
ratio  for  the  class  B tubes  to  be  used. 


656  Speech  and  A.  M.  Equipment 


THE  RADIO 


INTERSTAGE 
TRANSFORMER 
1 :3  STEP-UP 


304TL 


Figure  13 

PUSH-PULL  304TL  CLASS  AB,,  MODULATOR 


A three  wire  shielded  cable  should  be  used 
to  connect  the  6B4-G  tubes  to  the  driver  trans- 
former located  at  the  grids  of  the  class  B 
tubes.  This  cable  may  be  any  reasonable 
length  up  to  25  or  30  feet.  Any  of  the  modula- 
tor configurations  shown  in  figure  8 may  be 
driven  with  this  simple  speech  amplifier. 

30-5  500-Watt 

304TL  Modulator 

Ordinarily,  few  amateurs  would  be  inclined 
to  design  the  high-level  stages  of  their  trans- 
mitters around  type  304TL  tubes.  Although  the 
304TL  unquestionably  is  an  excellent  tube, 
with  its  extremely  high  transconductance  and 
power  sensitivity,  its  price  and  capabilities 
are  in  excess  of  those  required  even  for  a kilo- 
watt amateur  transmitter.  But  with  enormous 
supplies  of  these  tubes  available  at  a relative- 
ly low  price  on  the  surplus  market  it  becomes 
economical  to  consider  their  use  in  high-power 


TABLE  I 


304TL  OPERATING  CHARACTERISTICS,  CLASS  ABI 

( NO  DRIVING  POWER  REQUIRED  ) 

D.C.  PLATE  VOLTS 

1500 

2000 

2500 

3000 

GRID  BIAS  X- 

-170 

-230 

-290 

ZERO-SIGNAL 
PLATE  CURRENT 
{MA.) 

270 

200 

160 

130 

MAX.  SIGNAL 
PLATE  CURRENT 
(MA.)  X-  X 

572 

546 

463 

444 

PLATE-TO-PLATE 
LOAD  {.OHMS) 

2540 

5300 

8500 

12000 

PEAK  A.F. 

GRID  VOLTS 
(.PER  TUBE  ) 

1 18 

170 

230 

290 

MAX.  POWER 
OUTPUT  (WATTS) 

256 

490 

<10 

730 

X-  ADJUST  TO  GIVE  STATED  ZERO-SIGNAL  PLATE  CURRENT 

X-  VALUES  GIVEN  FOR  SINE-WAVE  MODULATION.  FOR  VOICE 
WAVE  FORMS.MAXIMUM  CURRENT  WILL  BE  APPROX  IMATELV 
TWO-THIRDS  SINE-WAVE  VALUES. 

Figure  1 4 


amateur  transmitters.  In  fact,  with  the  use  of 
these  304TL's  the  cost  of  heating  power  for 
filaments  becomes  a much  more  significant 
figure  than  the  cost  of  the  tubes. 

The  304TL  is  ideally  suited  for  use  as  a 
modulator  for  a high-power  amateur  transmit- 
ter. In  fact,  due  to  the  relatively  low  amplifi- 
cation factor  of  the  tube  (about  12)  the 
304TL  may  even  be  used  as  a class  A triode 
Heising  modulator.  Such  an  arrangement  is 
practicable  for  modulating  a medium-power 
transmitter  when  a 1500  to  2000  volt  plate 
supply  is  available. 

The  class  ABi  operating  characteristics  of 
the  304TL  tube  are  shown  in  figure  14.  An  in- 
teresting fact  to  be  observed  is  that  under 
these  conditions  no  grid  driving  power  is  re- 
quired by  the  304TL  tubes.  It  is  necessary 
only  to  supply  the  correct  amount  of  grid  driv- 
ing voltage.  The  amounts  shown  may  easily 
be  obtained  from  a single  6J5  tube,  and  the 
usual  push-pull  triode  driver  stage  for  the  high 
powered  modulator  may  be  eliminated.  A suit- 
able modulator-driver  unit  which  will  deliver 
500  watts  of  audio  from  two  304TL  tubes  op- 
erating class  ABi  is  shown  in  figure  13. 

The  filaments  of  the  304TL  tubes  may  be 
connected  either  in  series  or  parallel  for  either 
5 volt,  50  ampere  operation,  or  10  volt,  25  am- 
pere operation.  To  conserve  filament  power 
during  standby  periods,  the  filaments  of  the 
304TL  tubes  may  be  either  turned  off,  or  may 
be  dropped  to  one-half  voltage  by  means  of  a 
dropping  resistor  in  the  primary  circuit  of  the 
filament  transformer.  This  resistor  may  be 
shorted  out  during  periods  of  transmission  by 
a small  relay  actuated  by  the  transmitting  con- 
trol system. 

The  gain  of  the  speech  amplifier  is  suffi- 
cient so  that  an  inexpensive  crystal  micro- 
phone may  be  used  with  the  modulator.  The 


HANDBOOK 


15-Watt  Clipper-Amplifier  657 


12AX7  6AL5  12AU7  6B4-G  Jz 


"phone-c.w.”  switch  is  connected  so  as  to 
short  the  modulation  transformer  and  to  open 
the  filament  circuit  of  the  304TL  tubes  during 
c-w  operation. 

30-6  A 15-Wott 

Clipper  - Amplifier 

The  near-ultimate  in  "talk  power"  can  be 
obtained  with  low  level  clipping  and  filtering 
combined  with  high  level  filtering.  Such  a 
modulation  system  will  have  real  "punch," 
yet  will  sound  well  rounded  and  normal.  The 
speech  amplifier  described  in  this  section 
makes  use  of  low  level  clipping  and  filtering 
and  is  specifically  designed  to  drive  a pair  of 
push-pull  810  modulators  such  as  shown  in 
Section  Three. 

Circuit  The  schematic  of  the  speech 

Description  amplifier-clipper  is  shown  in 

figure  15.  A total  of  six  tubes, 
including  a rectifier  are  employed  and  the  unit 
delivers  1 5 watts  of  heavily  clipped  audio. 

A 12AX7  tube  is  used  as  a two  stage  mi- 
crophone pre-amplifier  and  delivers  approxi- 
mately 20  volts  (r.m.s.)  audio  signal  to  the 
6AL5  series  clipper  tube.  The  clipping  level 
is  adjustable  between  0 db  and  15  db  by  clip- 
ping control,  R=.  Amplifier  gain  is  controlled 


by  Ri,  in  the  grid  circuit  of  the  second  section 
of  the  12AX7.  A low  pass  filter  having  a 3500 
cycle  cut-off  follows  the  6AL5  clipper  stage, 
with  an  output  of  5 volts  peak  audio  signal 
under  maximum  clipping  conditions.  A double- 
triode  12AU7  cathode  follower  phase-inverter 
follows  the  clipper  stage  and  delivers  a 125 
volt  r.m.s.  signal  to  the  push-pull  grids  of  the 
6B4-G  audio  driver  tubes.  The  6B4-G  tubes 
operate  at  a plate  potential  of  330  volts  and 
have  a “68  volt  bias  voltage  developed  by  a 
small  selenium  rectifier  supply  applied  to  their 
grid  circuit.  An  audio  output  of  1 5 watts  is  de- 
veloped across  the  secondary  terminals  of  the 
class  B driver  transformer  with  less  than  5 per 
cent  distortion  under  conditions  of  no  clipping. 
A 5U4-G  and  a choke  input  filter  network  pro- 
vide unusually  good  voltage  regulation  of  the 
high  voltage  plate  supply. 

Amplifier  The  clipper-amplifier  may  be 

Consfruction  built  upon  the  same  chassis  as 
the  power  supply,  provided 
the  low  level  stages  of  the  amplifier  are  spaced 
away  from  the  power  transformers  and  filter 
chokes  of  the  supply.  All  capacitors  and  resis- 
tors of  the  audio  section  should  be  mounted  as 
close  to  the  respective  sockets  as  is  practical.  For 

minimum  hum  pickup,  the  filament  leads  to  the 
low  level  stages  should  be  run  in  shielded  braid. 


658  Speech  and  A.  M.  Equipment 


THE  RADIO 


Those  resistors  in  the  12AU7  phase  inverter 
plate  circuit  and  the  grid  circuit  of  the  6B4-G 
tubes  should  be  matched  to  achieve  best  phase 
inverter  balance.  The  exact  value  of  the  paired 
resistors  is  not  important,  but  care  should  be 
taken  that  the  values  are  equal.  Random  re- 
sistors may  be  matched  on  an  ohmmeter  to  find 
two  units  that  are  alike  in  value.  When  these 
matched  resistors  are  soldered  in  the  circuit, 
care  should  be  taken  that  the  heat  of  the  sol- 
dering iron  does  not  cause  the  resistors  to 
shift  value.  The  resistors  should  be  held  firm- 
ly by  the  lead  to  be  soldered  with  a long  nose 
pliers,  which  will  act  as  a heat-sink  between 
the  soldered  joint  and  the  body  of  the  resistor. 
If  this  precaution  is  taken  the  two  phase  in- 
verter outputs  will  be  in  close  balance. 

Adjustment  of  When  the  wiring  of  the 
the  Speech  speech  amplifier  has  been 

Amplifier  completed  and  checked,  the 

unit  is  ready  to  be  tested. 
Before  the  tubes  are  plugged  in  the  amplifier, 
the  bias  supply  should  be  energized  and  the 
voltage  across  the  600  ohm  bleeder  resistor 
should  be  measured.  It  should  be  ~68  volts. 
If  it  is  not,  slight  changes  in  the  value  of  the 
series  resistor,  Ra,  should  be  made  until  the 
correct  voltage  appears  across  the  bleeder  re- 
sistor. The  tubes  may  now  be  inserted  in  the 
amplifier  and  the  positive  and  cathode  voltages 
checked  in  accordance  with  the  measurements 
given  in  figure  15.  After  the  unit  has  been 


tested  and  is  connected  with  the  modulator, 
R2  should  be  set  so  that  it  is  impossible  to  over- 
modulate the  transmitter  regardless  of  the  set- 
ting of  Ri.  The  gain  control  ( Ri ) may  then  be 
adjusted  to  provide  the  desired  level  of  clip- 
ping consistent  with  the  setting  of  Ri. 

30-7  A 200-Wott 

811 -A  De-luxe  Modulator 

One  of  the  most  popular  medium  power  r-f 
amplifier  stages  consists  of  a single  tetrode 
tube,  such  as  the  4-125A,  813,  or  7094  oper- 
ating at  a plate  potential  of  1200  - 1700  volts 
and  a plate  current  of  150  - 275  milliamperes. 
Such  an  amplifier  requires  a minimum  of  r-f 
driving  power,  allows  an  input  of  300  to  400 
watts,  and  yet  employs  power  supply  com- 
ponents that  are  relatively  modest  in  cost.  The 
5-db  signal  increase  between  a 300  watt  trans- 
mitter and  a 1000  watt  transmitter  is  very 
expensive  when  one  considers  the  additional 
cost  of  modulator  and  power  supply  equipment. 

Additional  economy  may  be  achieved  if 
the  modulator  and  final  amplifier  are  operated 
from  the  same  power  supply.  The  new  series  of 
Chicago-Standard  plate  transformers  provide 
voltage  ranges  in  the  1000  to  1500  volt  re- 
gion and  are  well  suited  for  the  combination 
of  this  modulator  and  the  aforementioned  r-f 
tubes.  Within  this  voltage  range,  the  811-A 
triode  is  an  excellent  choice  for  the  modulator 
tubes.  Zero  bias  operation  may  be  had  up  to 


Figure  16 
REAR  VIEW  OF 
Sn-A  MODULATOR 

Modulator  tubes  and  voltage 
regulator  are  at  right  with 
high  level  filter  at  center  of 
chassis.  Plug-in  speech  am- 
plifier is  to  left  of  clipper. 
Gain  and  clipping  controls 
are  atop  the  small  chassis. 
6L6/S881  is  used  as  cathode 
follower  driver  stage  for 
modulator. 


HANDBOOK 


8 11 -A  Modulator  659 


Figure  17 

SCHEMATIC,  200  WATT  811-A  MODULATOR 

Ti — J;3  Interstage  Transformer,  Stancor  A-53 

Tt — "Poty-pedance"  class  B driver  transformer,  2:1  ratio,  Stancor  A-47S1 

T: — 200  watt  modulation  transformer,  9 K primary,  SK  secondary,  Stancor  A-3829 

T: — 400  - 0 - 400  volts,  250  ma,,  6.3  volts,  5 amperes,  Stancor  PC-8413 

Ts — 6.3  volts,  10  amperes,  Stancor  P-6308 

CHi — 4 henry,  2S0  ma,  Stancor  C-74I2 

Lj — Low  pass  filter,  3000  cycle  cut  off,  Chicago  LPF-2, 

Lr — "Splatter^'  filter,  300  ma,  Stancor  C-23I7 

RYt — SPOT  relay,  high  voltage  insulation,  115  volt  coil,  Leach  #7723  with  374  coils,  or  equivalent 


1250  volts,  and  only  4.5  volts  is  required 
for  1500  volt  operation.  Bias  voltage  may  be 
obtained  from  flashlight  batteries  or  other  low 
impedance  source. 

Modularor  The  200  watt  de-luxe  modulator 
Circuit  is  illustrated  in  figures  16  and 
18  and  the  schematic  is  given 
in  figure  17.  The  low  level  audio  stages  are 
similar  to  those  of  the  speech  amplifier  shown 
in  Section  Six.  A 12AX7  is  employed  as  two 


stages  of  R-C  amplification  driving  a 6AL5 
speech  clipper  tube.  A 3500  cycle  low  pass 
filter  follows  the  clipper,  removing  all  high 
order  products  of  clipping  action.  A parallel- 
connected  12AU7  follows  the  filter  and  is 
transformer-coupled  to  a 5881  (6L6-GB)  ca- 
thode follower  driver  stage.  The  impedance  of 
the  cathode  circuit  of  the  driver  stage  is  ex- 
tremely low;  it  provides  an  excellent  driving 
source  for  the  class  B modulator  grid  circuit. 


660  Speech  and  A.  M.  Equipment 


THE  RADIO 


Two  811-A  tubes  are  employed  in  the  class 
B stage.  When  operated  at  1000  volts,  no  bias 
supply  is  needed.  At  voltages  of  1200  or  above, 
approximately  9 volts  of  bias  is  required.  This 
is  supplied  by  a voltage  divider  composed  of 
a 20K,  10  watt  resistor  and  a 2D21  thyratron 
tube.  When  the  miniature  2D21  is  connected 
as  a triode,  it  acts  as  a voltage  regulator  tube 
with  a constant  voltage  drop  of  almost  9 volts 
from  plate  to  cathode.  The  tube  will  regulate 
over  300  milliamperes  of  current  while  main- 
taining a reasonably  constant  voltage  drop 
across  its  terminals.  The  center  tap  of  the 
811-A  filament  transformer  fTs)  is  thus  held 
at  a positive  potential  with  respect  to  ground. 
Since  the  center  tap  of  the  811-A  driver  trans- 
former (Tu)  is  grounded,  the  modulator  tubes 
are  biased  at  a constant  negative  voltage  equal 


to  the  voltage  drop  across  the  2D21  regulator 
tube  in  the  cathode  circuit  of  the  class  B stage. 

The  plate  to  plate  load  impedance  of  the 
811-A  tubes  when  operating  at  1500  volts  is 
approximately  12,000  ohms.  A multi-match 
type  modulation  transformer  may  be  employed 
if  desired,  but  in  this  case  a Stancor  A-3829 
unit  was  used.  This  transformer  is  designed  to 
match  the  plate-to-plate  load  impedance  of 
9,000  ohms  to  a secondary  load  of  5000  ohms. 
With  the  12,000  ohm  load  of  the  811-A  tubes, 
a secondary  load  of  7,500  ohms  must  be  used 
to  maintain  the  same  primary  to  secondary  im- 
pedance ratio.  This  secondary  load  can  be  ob- 
tained with  a single  7094  tube  operating  at 
1500  volts  and  200  milliamperes  of  plate  cur- 
rent (300  watts  input).  Other  tubes  and  load 
impedances  can  also  be  used,  providing  the 
r-f  input  to  the  modulated  stage  does  not  ex- 
ceed 400  watts.  For  example,  a 4-125A  tube 
operating  at  2000  volts  and  165  ma.  (330 
watts)  may  be  modulated  by  this  audio  unit. 
The  secondary  winding  of  the  modulation 
transformer  can  pass  a maximum  of  300  mil- 
liamperes with  safety. 

The  audio  output  from  the  811-A  stage  is 
passed  through  a high  level  low-pass  "splatter 
suppressor’’  which  attenuates  all  audio  fre- 
quencies above  3500  cycles.  The  use  of  both 
low  level  and  high  level  audio  filters  does 
much  to  reduce  the  broad  sidebands  and  co- 
channel interference  that  seems  to  be  so  com- 
mon on  the  amateur  phone  bands. 

A high  voltage  relay  RYi  is  employed  to 
short  the  secondary  of  the  modulation  trans- 
former and  remove  plate  potential  from  the 
modulator  tubes  for  c-w  operation.  The  relay 
is  actuated  by  the  "phone-c.w.”  switch  on  the 
front  panel  of  the  modulator.  Other  segments 
of  this  switch  turn  off  the  modulator  fila- 
ments and  provide  extra  contacts  to  control 
auxiliary  equipment. 

A 350  volt  supply  is  incorporated  in  the 
modulator  unit  to  power  the  speech  amplifier 
and  driver  stage  and  to  provide  power  for 
the  r-f  exciter  stages  of  the  transmitter.  The 
various  power  and  control  leads  are  brought 
out  to  a multi-connector  plug  mounted  on  the 
rear  of  the  modulator  deck. 


Figure  18 

UNDER-CHASSIS  VIEW  OF 
811 -A  MODULATOR 

High  voltage  relay  is  between  811 -A  tube 
sockets,  and  low  voltage  components  are  at 
opposite  end  of  chassis. 


HANDBOOK 


811 -A  Modulator  661 


Modulator  The  modulator  is  constructed 

Construction  upon  a steel  chassis  measur- 

ing 8"  X 17"  X 2".  A IOV2'' 
aluminum  panel  is  bolted  to  the  chassis  with 
two  mounting  brackets  to  form  a rugged  as- 
sembly. Placement  of  the  major  parts  may  be 
seen  in  figures  16  and  18.  The  modulation 
transformer  Ts  and  the  811 -A  tubes  occupy 
the  right  end  of  the  chassis,  balanced  in  weight 
by  the  power  transformer  T»  and  modulator 
filament  transformer  Ts  at  the  opposite  end  of 
the  chassis.  The  center  space  is  taken  by  the 
plug-in  speech  amplifier,  the  high  level  splat- 
ter filter  assembly  and  the  5881  driver  stage. 

The  speech  amplifier  is  constructed  as  a sep- 
arate unit  on  a small  aluminum  utility  box 
measuring  5"  x 3"  x 2".  The  bottom  of  the 
box  holds  two  male  plugs  which  match  two 
receptacles  mounted  on  the  amplifier  chassis. 
The  speech  amplifier,  therefore,  may  be  wired 
and  tested  as  a separate  unit.  Clipping  and 
audio  level  controls  are  mounted  atop  the 
amplifier  box  as  long  usage  of  clipper  circuits 
has  proven  that  these  controls  need  not  be  re- 
adjusted once  they  are  properly  set. 

The  phone-c.w.  switch,  relay  RY-i  and  var- 
ious small  components  are  mounted  beneath 
the  chassis  (figure  18).  The  input  receptacle 
for  the  speech  amplifier  box  is  located  adja- 
cent to  the  microphone  receptacle  on  the  front 
panel  of  the  modulator  making  the  intercon- 
necting lead  less  than  two  inches  long.  Also 
placed  beneath  the  chassis  are  the  filter  choke 
for  the  low  voltage  supply  and  the  various  by- 
pass and  filter  capacitors. 

Wiring  and  Testing  The  speech  amplifier 
the  Modulator  should  be  wired  first. 

The  small  resistors  and 
capacitors  are  mounted  either  between  the  tube 
socket  pins,  or  between  terminals  of  small 
phenolic  tie-point  strips.  Transformer  Ti  is 
fastened  within  the  amplifier  box  and  is  wired 
in  the  circuit  after  all  other  wiring  is  com- 
pleted. Plugs  PLi  and  PL2  are  mounted  on  the 
bottom  portion  of  the  box;  the  plug  pins  are 
wired  to  the  proper  points  of  the  speech  am- 
plifier with  short  lengths  of  wire  that  allow 
the  bottom  plate  to  be  removed  for  inspection 
and  testing  without  the  necessity  of  unsolder- 
ing any  connections  to  the  plugs. 

The  modulator  chassis  should  be  wired  next. 
All  leads  to  Ts,  RY-j,  and  the  low  pass  filter 
should  be  carefully  insulated  from  the  chassis. 
High  voltage  "5000  volt  test”  cable  should  be 
employed  for  these  connections.  The  capaci- 
tors that  make  up  the  high  level  audio  filter 
are  mounted  directly  to  the  terminals  of  the 


filter  choke  which  is  mounted  above  the  chas- 
sis on  hi -inch  ceramic  insulators.  High  voltage 
connections  to  the  modulator  are  made  through 
Millen  37001  safety  terminals. 

When  the  wiring  has  been  completed  and 
checked,  the  12AX7,  6AL5,  12AU7,  5881, 
and  5V4-GB  tubes  should  be  inserted  in  their 
sockets  and  the  speech  amplifier  is  plugged 
into  the  modulator  receptacles.  The  vertical 
amplifier  of  an  oscilloscope  should  be  connect- 
ed to  one  grid  terminal  of  the  811-A  stage. 
Plate  voltage  of  the  5881  should  be  approxi- 
mately 370  volts.  A low  level  1000  cycle  tone 
(approximately  0.05  volts,  r.m.s.)  is  applied 
to  the  amplifier  input.  The  output  level  of  the 
speech  amplifier  is  controlled  by  the  setting 
of  the  clipping  control  R-  and  the  audio  gain 
is  controlled  by  potentiometer  Ri  in  the  grid 
circuit  of  the  12AX7.  The  clipping  control 
should  be  set  so  that  not  more  than  60  volts 
r.m.s.  is  developed  from  one  811-A  grid  to 
ground.  The  modulator  tubes  may  now  be 
plugged  in  their  sockets.  A 7K,  200  watt  re- 
sistor should  be  placed  between  the  “H.V. 
Out”  and  "H.V.  In”  terminals,  serving  as  a 
dummy  load,  and  1500  volts  applied  to  the 
latter  terminal.  With  no  audio  signal  the 
resting  plate  current  of  the  modulator  stage 
should  be  approximately  15  milliamperes, 
kicking  up  to  about  I6O  milliamperes  under 
full  output  conditions.  Final  adjustment  of  the 
clipping  control  may  be  made  when  the  mod- 
ulator is  placed  in  use  with  the  r-f  section  of 
the  transmitter.  Potentiometer  R;  is  then  ad- 
justed to  limit  the  peak  modulation  level  under 
sine  wave  modulation  to  approximately  90%. 


Figure  19 

ZERO  BIAS  TETRODE  MODULATOR 
ELIMINATES  SCREEN  AND  BIAS 
SUPPLIES 

Low  driving  power  and  simplicity  are  key 
features  of  this  novel  modulator.  Tubes  rang- 
ing in  size  from  6AQ5*s  to  8l3's  may  be 
employed  in  this  circuit. 

Ti — Class  B driver  transformer 

Ti — Modulation  transformer 

Vj,  Vz — 6AQ5,  6L6,  807,  803,  813,  etc. 

Ri,  Rs — Not  used  with  803  and  813 


662  Speech  and  A.  M.  Equipment 


803's 


115  V.  'X, 


Figure  20 

INEXPENSIVE  500  WATT 
MODULATOR  USING  803  TUBES 

Ti — "Poly-pedanee"  Class  B driver  transform- 
er 2:1  ratio,  Stancor  A-4761 
T$ — 500  watt  output  transformer.  18K  primary, 
6.25K  secondary.  Chicago  CMSS 
T3 — 70  volts,  70  amperes.  Sioncor  C-6461 
M — 0 - 500  ma. 


30-8  Zero  Bios 

Tetrode  Modulators 

Class  B zero  bias  operation  of  tetrode  tubes 
is  made  possible  by  the  application  of  the 
driving  signal  to  the  two  grids  of  the  tubes  as 
shown  in  figure  19.  Tubes  such  as  the  6AQ5, 
6L6,  807,  803,  and  813  work  well  in  this 
circuit  and  neither  a screen  supply  nor  a bias 


supply  is  required.  The  drive  requirements 
are  low  and  the  tubes  operate  with  excellent 
plate  circuit  efficiency.  The  series  grid  resis- 
tors for  the  small  tubes  are  required  to  balance 
the  current  drawn  by  the  two  grids,  but  are  not 
needed  in  the  case  of  the  803  and  813  tubes. 

Of  great  interest  to  the  amateur  is  the  cir- 
cuit of  figure  20,  wherein  803  tubes  are  used 
as  high  level  modulators.  These  tubes  will  de- 
liver 500  watts  of  audio  in  this  configuration, 
yet  they  require  no  screen  or  bias  supply,  and 
can  be  driven  by  an  8 watt  amplifier  stage. 
The  use  of  803  tubes  (in  contrast  to  813’s) 
requires  a higher  level  of  driving  power  which 
is  offset  by  the  fact  that  these  tubes  can  often 
be  purchased  "surplus”  for  less  than  four  dol- 
lars. A pair  of  6B4  tubes  operating  with  ca- 
thode bias  ( figure  12)  will  suffice  as  a driv- 
ing stage  for  the  803’s.  The  power  supply  of 
the  speech  amplifier  provides  high  voltage  for 
the  suppressors  of  the  modulator  stage. 

Shown  in  figure  21  is  a high  level  modu- 
lator using  813  tubes.  A full  500  watts  of 
audio  may  be  obtained  at  2500  volts  plate  po- 
tential. Grid  driving  power  is  5.5  watts.  A 
single  807  operating  as  a cathode  follower 
at  400  volts  will  provide  sufficient  drive  for 
the  modulator  stage.  Plate  to  plate  load  im- 
pedance for  the  813’s  is  not  critical.  The  Chi- 
cago CMS-3  500  watt  modulation  transformer 
having  a primary  impedance  of  18,000  ohms 
has  been  used  with  success,  although  the  opti- 
mum plate  load  impedance  of  the  modulator 
is  closer  to  20,000  ohms. 


807 


DRIVER  STAGE,  OPERATING  VOLTS, 
MEASURED  TO  GROUND, 


807 

PIN  2 

300 

PIN3 

0 

PIN  4 

26.5 

NOTES 

^-£Jf/9CT  VALUE  OF  807  CATHODE  RESISTOR 
DEPENDS  UPON  RESISTANCE  OF  PRIMARY 
WINDING  OF  J2..  ADJUST  RESISTOR  FOR 
CATHODE  BIAS  OF  £6.5  VOLTS,  PLATE  CUR- 
RENT OF  S3~  55  MA . 

Z-RMS  OPERATING  VOLTAGES  AT  MAXIMUM 
OUTPUT  SHOWN  ON  SCHEMATIC. 


n 5 V, 


Figure  21 

500  WATT  MODULATOR  USING  813  TUBES 

Ti — 7:3  interstage  transformer.  Stancor  A-S3 

Tz — "Poiy-pedance"  Class  B driver  transformer.  2:1  ratio.  Stancor  A-4761 
Tj — 500  watt  output  transformer.  18K  primary,  6.2SK  secondary.  Chicago  CMS-3 
T, — 70  volts,  10  amperes,  Stancor  C-6461 
M — 0 - 500  ma. 

350  wafts  of  audio  are  obtainable  from  this  circuit  at  plate  potential  of  2000  volts. 


CHAPTER  THIRTY-ONE 


Despite  the  fact  that  complete  transmitters 
may  be  purchased  at  moderate  cost  many  ama- 
teurs still  prefer  to  build  their  station  equip- 
ment. The  experience  gained  in  constructing 
such  equipment  cannot  be  duplicated  by  any 
other  means. 

Shown  in  this  chapter  are  two  complete 
phone  and  c.w.  transmitters  that  are  well  suit- 
ed as  "building  projects”  for  the  advanced 
amateur  who  has  had  experience  building  other 
equipment.  The  transmitter  designs  are  unique 
in  that  equipment  of  this  type  and  power 
rating  is  not  readily  available  on  the  commer- 
cial market.  The  first  transmitter  covers  the  50 
Me.  and  144  Me.  VHF  bands  and  is  capable 
of  300  watts  input  (phone)  and  450  watts 
input  ( c.w. ) . The  second  transmitter  covers 
the  3.5  Me. — 29.7  Me.  amateur  bands  and  runs 
375  watts  input  (phone)  and  500  watts  input 
(c.w.).  Both  transmitters  are  complete,  includ- 
ing modulators  and  power  supplies  and  are 
system-designed  for  ease  of  operation  and  re- 
liability. The  cost  of  either  unit  is  quite  low, 
considering  the  power  level  involved. 


31-1  A 300  Watt 

Phone/C-W  Transmitter 
for  50/144  Me. 

Transmitters  operating  in  the  VHF  region  pre- 
sent some  specialized  design  problems  that  must 
be  solved  before  reliable,  efficient  operation 
can  be  obtained.  The  first  problem  concerns 
the  relative  inefficiency  of  tuned  and  coupled 
circuits  in  this  region.  The  amount  of  r-f  power 
that  can  be  lost  in  simple  circuits  is  extremely 
high  unless  special  pains  are  taken  in  the  design 
and  layout  of  the  VHF  stages.  Radiation  loss, 
skin  effect,  dielectric  losses,  lead  inductance, 
distributed  capacity,  and  parasitic  ground  re- 
turns all  exist  at  the  lower  frequencies,  but  only 
in  the  higher  portion  of  the  spectrum  do  their 
effects  assume  such  vital  importance,  and  only 
at  these  frequencies  does  their  cure  and  elimin- 
ation become  so  essential.  As  a result,  the 

necessity  for  close  correlation  between  the  elec- 
trical design  and  the  mechanical  properties  of 
the  VHF  circuit  cannot  be  overemphasized. 


663 


Figure  1 

COMPACT 
300  WATT 
TRANSMITTER  FOR 
SIX  AND  TWO 
METERS  IS  IDEAL 
FOR  VHF 
ENTHUSIAST 

The  transmitter  is  built 
in  two  sections.  Upper 
deck  contains  VFO,  r.i. 
stages  and  push-pull  826 
amplifier.  VFO  tuning 
dial  is  at  right,  with 
bandswitch  and  meter 
selector  switch  below  it. 
At  center  are  the  multi- 
plier tuning  controls, 
with  buffer  and  final 
amplifier  tuning  controls 
at  the  left. 

Lower  deck  holds  modu- 
lator and  power  supply. 
Transmitter  control 
switches  are  along  lower 
edge  of  the  panel. 


The  second  problem  concerns  the  vacuum 
tube.  A combination  of  factors  is  at  work  in 
the  common  tube  which  tend  to  limit  the  ef- 
ficiency of  the  tube  as  the  frequency  of  opera- 
tion is  raised.  Inter-electrode  capacitance,  lead 
inductance,  transit  time,  and  input  circuit  load- 
ing are  some  of  the  factors  that  cripple  va- 
cuum tube  performance  in  the  VHF  region. 
The  drop  in  mbe  operation  efficiency  shows 
up  in  the  form  of  plate  dissipation.  Reducing 
the  size  of  the  vacuum  tube  to  combat  the 
various  operational  problems  reduces  its  power 
handling  capability,  since  only  small  element 
areas  are  then  present  for  heat  dissipation. 

Tetrode  vs.  The  combination  of  intet-elec- 

Triode  trode  capacitance  and  lead  in- 

ductance in  the  simple  tetrode 
tube  produce  a phenomenon  called  self-neu- 
tralization. At  some  one  particular  frequency 
in  or  near  the  VHF  region  the  tetrode  tube  is 
inherently  self-neutralized  due  to  the  circuit 
elements  within  the  tube  structure  and  the  ex- 
ternal inductance  of  the  screen-to-ground  path. 
Below  this  frequency  normal  neutralizing  pro- 
cedures apply;  but  above  this  frequency  special 
neutralizing  circuits  are  required.  These  circuits 
ate  usually  frequency  sensitive.  The  self-neu- 
tralizing frequency  of  the  less  expensive  screen 
grid  tubes  usually  falls  between  20  - 50  Me., 
requiring  that  the  tubes  be  re-neutralized  when 
the  frequency  of  operation  is  changed  from 
50  Me.  to  144  Me.,  and  often  when  the  oper- 
ating frequency  is  moved  about  within  one 
of  these  bands.  The  newer  external  anode 
tetrodes  are  free  of  this  problem  as  their  self- 
neutralization frequency  falls  well  above  the 
144  Me.  amateur  band. 


The  triode  tube,  on  the  other  hand,  has  no 
frequency  of  self-neutralization.  The  maximum 
frequency  at  which  neutralization  is  effective 
is  dependent  upon  the  inductance  of  the  tube 
leads,  the  electron  transit  time,  and  the  inter- 
electrode  capacitance.  Rather  than  a point  in 
the  spectrum  at  which  the  neutralization  prob- 
lems abruptly  change,  the  neutralization  point 
of  the  triode  tube  merely  grows  more  obscure 
as  the  operating  frequency  is  raised,  until 
above  a certain  frequency  neutralization  is  no 
longer  possible.  Fortunately,  small,  compact, 
inexpensive  triode  tubes  exist  that  are  capable 
of  efficient  operation  well  into  the  VHF  re- 
gion, and  two  tubes  of  this  general  type  are 
employed  in  this  transmitter.  Special  circuits 
are  also  used  to  ensure  that  the  tubes  remain  in 
neutralization  over  the  operating  range. 

Transmitter  A block  diagram  of  the  r-f 
Circuitry  circuitry  is  shown  in  figure  2. 

The  transmitter  is  divided  into 
two  sections.  The  exciter  section  (Vi-V,)  de- 
livers about  ten  watts  of  power  over  the  fre- 
quency ranges  of  50-  54  Me.  and  144-148 
Me.  A 5763  is  employed  either  as  a crystal 
controlled  oscillator  or  as  a variable  frequency 
oscillator,  operating  in  the  8-9  Me.  region. 
Two  separate  oscillator  tuned  circuits  are  em- 
ployed as  the  tuning  rates  for  the  two  bands 
are  different.  The  "2  meter”  VFO  covers  the 
range  of  8.0  - 8.222  Me.,  and  the  "6  meter” 
VFO  tunes  8.333  - 9.0  Me.  The  tuning  capaci- 
tors of  the  two  circuits  are  ganged  for  tuning 
ease.  Tubes  V2  and  Vs  multiply  the  overall 
VFO  range  to  the  48  - 54  Me.  spectrum.  At 
this  point  a simple  switching  system  permits 
Vi  to  drive  either  the  intermediate  amplifier 


HANDBOOK 


50/144  Me.  Transmitter  665 


V4 

6360  {U4-U6  MC  ) 


© 


Figure  2 

BLOCK  DIAGRAM  OF  VHF  TRANSMITTER 

A — General  transmiiter  circuity  showing  operating  frequencies  of  the  various  stages,  V4  is  a plug-in 
frequency  multiplier. 

B — Quarter-wave  tank  circuit  is  employed  in  amplifier  plate  configuration  at  144  Me.  Shorting  bar 
is  removed  for  SO  Me.,  and  Cs-Ltt  resonates  at  this  frequency. 

C — Linear  half-wave  tank  section  is  employed  as  interstage  coupler  at  144  Me.  Inductor  Lio  resonates 
circuit  to  SO  Me. 


tube  Vs  directly  for  50-  54  Me.  operation,  or 
via  a tripler  stage  (Vi)  for  144-148  Me. 
operation.  A 9903/5894  (Vs)  is  employed 
as  the  driver  stage  for  the  push-pull  triode 
final  amplifier.  This  driver  tube  operates  as  a 
straight  amplifier  on  both  bands. 

A unique  coupling  circuit,  shown  in  figure 
2C,  is  employed  between  the  9903/5894  and 
the  push-pull  826  stage.  The  plates  of  the 
buffer  and  grids  of  the  final  stage  are  coupled 
to  the  ends  of  a loaded,  half-wave  144  Me. 
transmission  line.  The  50  Me.  coil  is  placed 
at  the  electrical  center  of  the  144  Me.  line. 
Tuning  capacitor  C?  is  tapped  on  the  line  at  a 
point  which  permits  each  circuit  to  be  resonat- 
ed. Thus  the  combination  LsA-B,  L»,  Ct  is 
capable  of  tuning  the  2 meter  and  6 meter 
ranges  simultaneously,  eliminating  the  need 
for  plug-in  coils  or  coil  switching.  Neutraliza- 
tion of  the  9903/5894  is  not  required  if  ade- 
quate inter-stage  shielding  is  used. 


The  power  amplifier  stage  employs  two  type 
826  tubes  in  push-pull.  These  compact  triode 
tubes  are  well  suited  to  VHF  work  as  their 
lead  inductance  is  small  and  their  inter-elec- 
trode capacitance  is  reasonably  low.  The  tubes 
have  adequate  plate  dissipation  and  a high  re- 
serve of  filament  emission.  The  amplifier  has 
been  run  at  one  kilowatt  input  (2000  volts 
at  500  milliamperes ) for  c.w.  operation  for 
extended  periods  of  time  without  apparent 
damage  to  the  tubes,  but  this  practice  is  not 
recommended  for  maximum  tube  life. 

A modified  form  of  two  band  plate  .circuit 
is  employed  in  the  amplifier  stage,  as  shown 
in  figure  2B.  A quarter-wave  144  Me.  trans- 
mission line  is  employed,  tuned  to  frequency 
by  "butter-fly”  capacitor  Cs.  The  "cold”  end 
of  the  line  is  shorted  by  a heavy  strap  which 
is  removed  for  50  Me.  operation.  At  the  latter 
frequency,  resonance  is  established  by  capaci- 
tor Cs  and  coil  L12  which  is  mounted  at  the 


?5  noofi 


Figure  3 

SCHEMATIC,  R-F  SECTION  OF  VHF  TRANSMITTER 


C,A-B — 25-25  niitd.  Bud  LC-I66I 
Cj,  Cs — 75  - 15  iJ.iJ.td.  Bud  LC-7660 
,,  Ci — 8 - 8 tijifd.  Bud  LC-16S9 
r— 70-  10  fififd.  ^'butterfly/'  Johnson  IIMBIl 
, — 35  - 35  fitifd.  Bud  LC-1667 
s — 25  - 25  /mfd.  "butterfly/'  B&W  JCX-2SB, 
stator  removed  in  each  section 


C<7— 35  Mfifd,,  ,024"  spacing.  Bud  LC-1872 

Cio — 50  fififd.,  .024"  spacing.  Bud  LC-1873 

Cii,  Cii — 1 - 7 fi/ifd.  See  figure  13 

RFCi — 2.5  m/i.  National  R-700 

RFCt — Ohmite  Z-50 

PFCi — Ohmite  Z-144 


SiA’B — 2 deck,  3 pole,  6 position  rotary  ceramic  switch. 

PLi — Octal  plug,  Amphenol 

PLt — Noval  plug  (9  pin)  See  text 

PLa — Octal  plug,  Amphenol 

Li-Lie — See  figure  12,  coil  table 

Xi-X;— 8.0  - 9.0  Me.  crystals,  type  FT-243 


-I 


> 

D 

O 


666  Transmitter  Construction 


HANDBOOK 


50/144  Me.  Transmitter  667 


Figure  4 

REAR  VIEW  OF  VHF  TRANSMITTER 

Final  amplifier  compartment  is  at  right  with 
antenna  receptacles  and  loading  capacitors  on 
rear  of  enclosure.  Ventilation  holes  are  drilled 
above  826  tubes  to  aid  air  circulation.  Entire 
top  should  be  replaced  with  perforated  ma- 
terial if  extended  periods  of  operation  are 
encountered.  Buffer  compartment  is  to  left 
of  large  enclosure,  with  9903/5894  tube 
mounted  in  inverted  position.  Grid  coil  (Lg, 
see  figure  7)  plugs  in  holder  at  top  of  com- 
partment. Grid  tuning  capacitor  is  panel  driv- 
en by  phenolic  shaft  extension.  The  144  Me. 
multiplier  box  is  in  position  behind  buffer 
compartment 

Variable  frequency  oscillator  is  at  left  of 
chassis,  with  plug-in  crystals  mounted  on  the 
side.  Oscillator  tube  Vi  can  be  seen  in  front 
of  enclosure.  Low  voltage  power  supply  com- 
ponents are  in  left  corner  of  chassis. 


ual  cross-neutxalization  system  is  not  effective 
over  the  range  of  50  - 148  Me.  The  amplifier 
may  easily  be  neutralized  at  either  end  of  the 
spectrum,  but  the  correct  capacitor  settings  at 
50  Me.  are  not  the  same  as  the  settings  for 
144  Me.  In  either  case,  however,  one  setting 
of  the  capacitors  is  sufficient  for  the  stage  to 
remain  in  complete  neutralization  across  one 
amateur  band.  For  a time  the  amplifier  was 
re-neutralized  each  time  the  frequency  of  op- 
eration was  shifted  from  band  to  band,  but 
this  chore  was  finally  eliminated  by  the  use 
of  the  neutralization  system  shown  in  the  sche- 
matic of  figure  3-  Conventional  cross-neutrali- 
zation (capacitors  Cn,  Cu)  is  employed  at  144 


B-plus  end  of  the  transmission  line.  The  144 
Me.  linear  lines  (LnA-B)  serve  merely  as  con- 
nections between  the  tubes  and  the  coil  at  50 
Me.  An  attempt  was  made  to  do  away  with 
the  shorting  bar,  permitting  the  amplifier  tank 
circuit  to  operate  in  the  same  manner  as  the 
buffet  circuit  which  requires  no  component 
changes  when  the  operating  frequency  is 
changed  from  band  to  band.  It  was  found  that 
the  resulting  drop  in  plate  circuit  efficiency 
was  so  high  as  to  be  unacceptable.  Wing  nuts 
ate  therefore  provided  to  permit  rapid  re- 
moval of  the  shorting  bar  when  50  Me.  oper- 
ation is  desired. 

Amplifier  Because  of  electron  transit 

Neutralization  time,  and  the  combination  of 

Circuit  lead  inductance  and  inter- 

electrode capacitance  the  us- 


Figure  5 

CLOSE-UP  OF  INTERIOR  OF 
VHF  FINAL  AMPLIFIER 

Two  826  triode  tubes  are  used  in  unique  push- 
pull  circuit  for  50  Me.  and  144  Me.  operation. 
Linear  tank  circuit  is  employed  for  144  Me. 
with  removeable  shorting  bar  in  foreground. 
"Butterfly"  tuning  capacitor  is  mounted  on 
ceramic  insulators  above  linear  tank  and  con- 
nected to  it  by  means  of  copper  straps  on  rear 
stator  terminals.  50  Me.  inductor  is  placed  at 
"cold"  end  of  2-meter  tank. 

Air  is  drawn  into  enclosure  through  ventila- 
tion holes  drilled  below  826  sockets.  Note  that 
the  sockets  are  split  to  reduce  possibility  of 
flash-over. 

At  left  of  enclosure  are  6360  (V-3)  and 
sockets  SO- 7 and  50-2  for  144  Me.  plug-in 
frequency  multiplier.  The  SO  Me.  adapter 

PL-3  is  in  place.  Oscillator  tuning  slugs  ore 

seen  on  right  side  of  oscillator  compartment 
of  left.  50  Me.  coil  is  in  place  atop  buffer 
compartment. 


PS? 


Figure  6 

UNDER-CHASSIS 
VIEW  OF  300 
WATT  VHF 
TRANSMITTER 

Small  nS~volt  blower 
motors  for  826  tubes  are 
in  upper  left  corner  of 
chassis.  Buffer  plate  line 
LgA-B  occupies  left-hand 
portion  of  chassis.  All 
leads  are  kept  clear  of 
this  area.  Under-chassis 
power  leads  are  run  in 
shielded  braid  to  reduce 
r-f  pick-up.  An  ''U'- 
shaped  shield  partition 
isolates  frequency  doub- 
ler stages  from  buffer 
plate  line.  Capacitors  Cs 
and  Cs  are  ganged,  as 
are  capacitors  Cs  and  C/,. 
Power  supply  components 
occupy  lower  right-hand 
corner  of  chassis. 

The  top  of  the  9903/- 
5694  buffer  tube  pro- 
tects through  hole  at 
center  of  chassis. 


Me.,  and  a simple  link  neutralization  circuit 
(Li4-Lib)  is  employed  to  complete  neutrali- 
zation at  50  Me.  This  double  neutralization 
circuit  results  in  absolute  stability  of  opera- 
tion on  both  bands.  Since  the  50  Me.  plate 
coil  Li2  is  decoupled  from  the  144  Me.  tank 
circuit  by  the  shorting  bar,  very  little  spurious 
coupling  exists  between  the  two  neutralizing 
circuits  and  adjustments  are  simple  and  may 
be  made  once  and  then  forgotten. 

The  Band-  A frequency  shift  between 

changing  System  50  Mc.  and  144  Me.  may 
be  made  in  a matter  of 
minutes.  Switch  Si  selects  the  cortea  crystal  or 
VFO  range.  The  plate  coil  La  is  broadly  re- 
sonant over  the  8-9  Me.  range  and  requires 
no  adjustment  after  initial  setting.  The  5763 
and  6360  plate  circuits  may  be  resonated  to 
the  correct  frequency  for  either  band  without 
the  necessity  of  coil  changing.  For  144  Mc. 
operation,  however,  it  is  necessary  to  remove 
the  50  Mc.  plug,  PL-3  (figure  7)  and  insert 
in  its  place  the  144  Mc.  triplet  box  (figure  3). 
In  addition,  grid  coil  Ls  of  the  9903/5894 
stage  must  be  changed.  The  amplifier  plate 
circuit  shorting  strap  is  removed  for  144  Mc. 
Finally,  the  buffer  and  amplifier  plate  tank 
circuits  must  be  resonated  to  frequency. 

Separate  antenna  coupling  systems  ate  em- 
ployed for  each  band.  Two  coupling  loops  ate 
used,  the  144  Mc.  loop  (Lm)  is  coupled  to 
the  quarter- wave  tank  circuit,  and  the  50  Mc. 
loop  (Lm)  is  coupled  to  plate  coil  L12. 


Power  Supply  The  low  voltage  supplies  are 
and  Modulator  located  on  the  rear  of  the  r-f 
deck,  and  are  diagrammed  in 
figure  8.  A voltage  doubler  selenium  supply 
provides  600  volts  for  the  9903/5894  stage 
and  270  volts  for  the  exciter.  A second  sele- 
nium rectifier  supply  provides  negative  bias 
for  the  buffer  and  final  amplifier  stages.  Iso- 
lating diodes  Dg  and  D;  reduce  interaction  be- 
tween the  two  bias  voltages. 

The  high  voltage  plate  supply  and  modula- 
tor are  located  on  a separate  deck.  The  supply 
provides  1500  volts  at  500  ma.  for  the  modu- 
lator and  final  amplifier  stage.  The  modulator 
consists  of  a pair  of  811-A  tubes,  operating 


.‘i 

Figure  7 
CLOSE  UP  OF  PLUG-IN 
COMPONENTS 

744  Mc.  plug-in  tripler  box  containing  6360 
(V-4)  is  at  center.  Tube  is  mounted  at  angle 
on  small  plate  with  ventilation  holes  drilled 
in  box  above  top  of  tube.  9-pin  plug  PLg  is 
made  from  portion  of  a Vector  socket.  50  Mc. 
plug  PLs  is  at  upper  left,  with  144  Mc.  coil 
Ls  directly  below  it.  50  Mc.  coil  Ls  is  at  right. 

See  figure  12  for  coil  data. 


50,/144  Me.  Transmitter  669 


Figure  S 

SCHEMATIC,  LOW  VOLTAGE  SUPPLIES  AND  METER  SWITCHING  CIRCUIT 

Ti — 600/300  V.,  e.t.,  360  ma.  Use  300  rolt  taps  for  voltage  doubler  service.  Triad  P-3. 

Ti — 125  voits,  50  ma.  Stancor  PA-8421 
Ti — 7.5  volts,  8 a.  Stancor  P-6738 
Ti — 6.3  volts,  6 a.  Stancor  P-3036 

Di-s-i-i — Selenium  rectifier,  100  ma.,  RMS  input  voltage:  175.  Maximum  peak  inverse  voltage;  495. 
Sarkes-Tarzian  Model  108 

Ds-i-7 — Selenium  rectifier,  75  ma.,  RMS  input  voltage:  130.  Sarkes-Tarzian  Model  75 
Bi,  Bi — Small  llS-volt  blower  motor. 

Si — 2 pole,  6 position  rotary  switch. 

SOj — 8 prong  octal  socket  (see  figures  3 and  10) 


class  B.  Any  one  of  the  speech  amplifiers 
shown  in  chapter  30  is  suitable  for  use  with 
the  modulator.  The  complete  schematic  of  this 
deck  is  shown  in  figure  10. 

High  voltage  primary  circuits  are  energized 
by  relay  RYi  after  the  plate  circuit  control 
switch  S2  is  closed.  External  relay  control  ter- 
minals are  provided  for  auxiliary  circuits. 

Metering  A 0-100  d.c.  milliammeter  (M:, 
Circuits  figure  10)  is  employed  to  monitor 
the  low  voltage  circuitry.  A six 
position  switch  ( S2,  figure  8 ) places  the  meter 
across  150  ohm  shunts  in  the  plate  and  grid 
circuits  as  shown  in  the  table  of  figure  8. 
Meter  Mi  reads  plate  current  of  the  modulator 
and  final  r-f  amplifier.  This  meter  is  placed 
in  a high  potential  circuit,  and  should  have  a 
bakelite  case  and  an  insulated  zero-set  control. 

Transmitter  The  overall  layout  of  the  r-f 
Construction  section  can  be  seen  in  figures 
4,  5,  and  6.  The  foundation 


consists  of  a 13”  x 17”  x 3”  aluminum  chassis 
firmly  fastened  to  a 1214”  x 19”  relay  rack 
panel.  The  oscillator  compartment  (figure  9) 
measures  2"  x 4”  x 6".  The  grid  circuit  com- 
partment for  the  9903/5894  is  3”  x 4"  x 5” 
in  size,  and  the  plug-in  144  Me.  triplet  box 
measures  2”  x 3"  x 314”.  The  push-pull  826 
final  amplifier  is  enclosed  in  an  aluminum 
box  measuring  6”  x 7”  x 12”.  This  compart- 
ment is  mounted  clear  of  the  chassis  on  14- 
inch  metal  posts.  Care  must  be  taken  to  see 
that  each  corner  of  the  box  makes  a good  elec- 
trical ground  with  the  chassis.  The  plug-in 
triplet  box  fits  into  two  tube  sockets  (SO-1 
and  SO-2)  placed  in  line  with  the  extension 
shaft  that  drives  buffet  grid  capacitor  Ce.  The 
special  sockets  with  50  Me.  plug  PL- 3 inserted 
in  SO-1  can  be  seen  in  figure  5. 

The  9903/5894  buffer  socket  is  mounted  to 
the  "top”  of  the  grid  circuit  compartment  with 
the  plate  leads  projecting  out  of  matching  holes 
cut  in  the  "bottom”  of  the  box  and  the  chassis. 


670  Transmitter  Construction 


THE  RADIO 


Figure  9 
INTERIOR, 

HIGH  STABILITY 
VFO  FOR  VHF 
TRANSMITTER 

Two  gang  oscillator  tun- 
ing capacitor  CiA-B  is 
panel  mounted  at  top  of 
enclosure.  Variable  trim- 
ming capacitors  are  at 
the  right,  with  slug-tuned 
coils  directly  below. 
Bandswitch  5iA-B  is 
placed  at  bottom  of  front 
panel,  with  crystal  sock- 
ets at  left.  The  two  leads 
to  oscillator  tube  socket 
pass  through  large  rubber 
grommet  mounted  on 
chassis  deck. 

After  alignment  is  com- 
pleted, a drop  of  nail 
polish  is  placed  on  end 
of  coil  slug  to  eliminate 
mechanical  vibration  of 
slug.  Note  that  rear  of 
case  is  bolted  in  place 
with  eight  sheet  metal 
screws  to  provide  rigidity 
to  enclosure.  Oscillator 
tube  socket  Is  in  fore- 
ground, partially  hidden 
by  selenium  rectifier. 


Ventilation  holes  are  drilled  in  the  top  of  this 
box,  and  grid  coil  Ls  mounts  in  banana  plugs 
mounted  on  a polystyrene  plate  atop  the  box. 

Two  826  sockets  are  mounted  at  the  front 
end  of  the  amplifier  box  on  metal  spacers. 
Trouble  occurred  with  the  sockets  flashing 
over  on  occasional  modulation  peaks.  Accord- 
ingly, that  portion  of  the  socket  holding  the 
two  plate  pins  was  cut  away  and  the  pin  con- 
nectors were  bolted  directly  to  the  ends  of  the 
plate  circuit  straps,  LnA  and  LuB.  The  back 
of  a hacksaw  blade  and  cutting  compound 
are  used  to  cut  grooves  across  the  top  and 
bottom  surfaces  of  the  socket.  A sharp  blow 
with  a hammer  will  fracture  the  socket  along 
the  groove  lines. 

Ventilation  holes  are  drilled  in  the  box  and 
chassis  directly  below  the  tubes  and  are  covered 
with  window  screen  to  reduce  spurious  energy 
radiation  through  the  openings. 

The  two  IVs”  wide  aluminum  straps  that 
serve  as  the  144  Me.  tuned  circuit  are  support- 
ed at  the  "B-plus”  end  on  two  solid  ceramic 
insulators.  Two  more  insulators  of  the  same 
height  serve  to  mount  the  "butter-fly”  capaci- 
tor to  the  box.  Small  copper  strips  connect  the 
tank  straps  to  the  stator  terminals  of  the  vari- 
able capacitor.  The  50  Me.  coil  and  the  re- 


Figure  1 1 
REAR  VIEW, 

POWER  SUPPLY  CHASSIS 

Power  supply  is  built  upon  steel  chassis.  Power 
transformer  and  modulation  transformer  are 
at  rear  of  chassis.  Three  4 fifd.,  2KV  filter 
capacitors  are  connected  in  parallel  to  pro- 
vide adequate  dynamic  reserve.  One  capacitor 
is  placed  under  the  chassis.  Insulated,  high 
voltage  terminal  is  employed  for  B-plus  lead. 
Rectifiers  are  at  right,  front  of  chassis,  with 
modulator  tubes  at  left-front. 


HANDBOOK 


50/144  Me.  Transmitter  671 


Tz  866 


B+TO 

FINAL 

811  -A 

TO 



AMPLIFIER 

'lH-fo3^6oy^  SO  4 
LJ^2,V47c^.^  to  r.f,  deck,  S03 


S3  I 

MK  PLATE  I Or 


Si 

FILAMENT 


EXTERNAL 

Oj  RELAY  CONTROL 

Figure  1 0 

SCHEMATIC,  POWER  SUPPLY  AND  MODULATOR  DECK 

— 1775/1500  each  side  c.t.,  625  ma.,  /CAS.  IJ5/230  yoH  primary.  Stancor  P-8029 
— 2.5  V.,  10  a.,  10KV  insulaiion.  Stancor  P-3060 
— 6.3  Y.,  10  o.  Stancor  P-6308 

— 300  watt  "polypedance"'  modulation  transformer.  Stancor  A-3898.  Primary  impedonce  12K, 
secondary  impedance  6K 

— "Polypedance"  driYer  transformer.  Stancor  A-4762 
'i — DPDT  relay,  10  ampere  contacts,  115  y.  a.c.  coil 
— 0 - 500  ma. 

— 0 - 700  mo. 

\-B — 2 pole,  3 position  high  Yoltage  switcfi.  (Found  in  "'war  surplus"  0C-3O6A  antenna  tuning  unit), 
or  equiYalent 

h — 500  mo.  swinging  choke,  2-76  h.  Stoncor  C-7405 
•i-3 — 7 75  Yolt  pilot  lamps 
'j, — 8 prong  octal  socket 


movable  shorting  bar  are  mounted  at  the  "B- 
plus”  end  of  the  assembly.  The  rotor  of  the 
"butter-fly”  capacitor  is  left  floating  and  is 
panel  driven  through  a high  voltage  shaft 
coupling.  The  neutralizing  capacitors  (Gi,  G2) 
are  mounted  on  a triangular  piece  of  poly- 
styrene or  other  good  insulating  material  and 
placed  between  the  tube  sockets.  Construction 
of  these  capacitors  is  shown  in  figure  13. 

The  144  Me.  antenna  link  "coil”  is  a U- 
shaped  pickup  loop  mounted  beneath  the  linear 
tank  circuit,  while  the  50  Me.  pickup  coil  is 
inserted  between  the  turns  of  plate  coil  L12. 
The  single  turn  neutralizing  coil  Ln  is  placed 
near  one  end  of  L12  as  shown  in  the  photo- 
graph. 


Placement  of  the  major  components  beneath 
the  chassis  may  be  seen  in  figure  6.  The  linear 
tank  circuit  LgA-B  runs  in  zig-zag  fashion 
from  the  plate  terminals  of  the  buffer  tube  Vs 
to  two  ceramic  standoff  insulators  mounted  on 
the  bottom  plate  of  the  power  amplifier  en- 
closure. Spacing  between  the  lines  is  held  con- 
stant, and  they  are  supported  at  the  buffer 
end  on  two  1"  ceramic  insulators.  The  buffer 
tuning  capacitor  C7  is  mounted  to  the  chassis 
between  the  lines  and  panel  driven  with  a 
phenolic  shaft.  The  rotor  of  this  capacitor  is 
not  grounded.  The  stators  of  Ci  are  connected 
to  the  lines  with  short  lengths  of  flexible  strap. 

Any  grid  current  unbalance  in  the  final  am- 
plifier stage  can  be  corrected  by  varying  the 


672  Transmitter  Construction 


THE  RADIO 


FIGURE  12 

COIL  CHART  FOR  50-144  MC.  TRANSMITTER 


Ll  -{144  MC.  COIL  ) 36  T.#24  1/2"  DIA.  MILLEN  69046  FORM. 

L2-(50MC.  CO/L  ) 28T.#24E.,  1/2"DIA.  M/LLEA/  69046  FORM. 

L3-  52  TURNS  1*  28  e.  M/LIEAI  69046  FORM. 

L4-  22  T.  center  tap,  3/4"  Dl  A . , 32  T.  P.  I . 30  7 £ 

LS-  15  T.  CENTER  TAP,  1/2*  Dl  A.,  1 6 T.P.  I . B6 IV  3 003 
LINK  COIL;  ONE  TURN  HOOKUP  WIRE. 

Le-  2 1/2  T,  CENTER  TAP,  1/2"  Dl  A.  16  T.  P.  I , ff£ty  3003 

LINK  coil:  one  turn  HOOKUP  WIRE.  I 

L?-  7 T.  CENTER  TAP,  5/8*DlA.,  16T.P,I.  BA  IV  300?  I 

LINK  COI  L ; ONE  TURN  HOOKUP  WI  R E . ' 

Le-  {50MC.  ) IT.  CENTER  TAP,  3/4"  01  A.,  6 T.P.  I.  BAW  3070  I 
LINK  COIL  ; ONE  TURN  HOOKUP  WIRE.  I 

{744  MC.)3  1/4*  X .02*  X .125"  COP.  STRAP  BENT  ""^“.{SEE FIO.  7)1 
LINK:  1 TURN  HOOKUP  WIRE  BENT  TO  FIT  L6. 

LQA-B-  1 1 " »8  COPPER  WIRE,  OR  .125"  Dl  A.  ROD,  PLUS  3"  OF 
1/2*  WIDE  FLEXIBLE  COPPER  STRAP  {BRIO  LEAD)  PLUS 
3/4"  FLEXIBLE  COPPER  STRAP  [PLATE  LEAD  ).  ADJUST 
SPACING  FOR  RESONANCE.  6 FOR  LAYOUT) 

LiO-5T.,  3/4"DIA.,#14E.  WIRE.  ADJUST  SPACING  OF  TURNS 
FOR  RESONANCE  AT  50  MC.  WITH  C?  AT  FULL  CAPACITY. 

LiiA-B  - 10"  OF  1/8"  X 1 1/8"  ALUMINUM  STRAP,  SPACED  3 3/8"  . 
CENTER  TO  CENTER.  SHORTING  BAR  MADE  OF  SAME  MAT-  I 
ERIAL,  ATTACHED  6 7/8"  FROM  PLATE  END  TO  CENTER  I 
OF  BAR.  PLATE  PIN  RECEPTACLES  REMOVED  FROM  SOCKET  ' 
AND  ATTACHED  TO  ENDS  OF  LINES  [SEE  FIGURE  S ) 

Lia-fSOAfC,  COIL  ) 6 T.,  1 1/4  DIA.  #6  COPPER  WIRE.  OR  .125" 
TUBING.  SPACE  TURNS  TO  RESONATE  TO  50  MC.  WITH  i 
CSNEAR  FULL  CAPACITY. 

Li3-19*OF#12E,WIRE  BENT  TO  "U'  SHAPE  TO  FORM  3' X 2"  , 

LOOP  WITH  3 1/2"  LEADS.  PLACE  UNDER  LllA-B, 

Ll4-1  TURN,  2"  DIA.,  10  KV,  HOOKUP  WIRE  . 

Li5-1  TURN,  1/2"  DIA.  #16  £.,  WITH  INSULATED  SLEEVING.  j 
PLACE  AT  CENTER  OF  LlO.  ' 

Lie-  2 TURNS,  1 1/4"  DIA,,  lO  KV  HOOKUP  WIRE  . 


position  of  the  taps  on  the  lines  until  balanced 
current  is  obtained.  The  tuning  range  of  the 
capacitor  on  144  Me.  is  reduced  by  moving 
the  taps  closer  to  the  electrical  center  of  the 
line. 

An  "L”-shaped  shield  plate  passes  around 
the  edge  of  triplet  capacitor  C.-Cs  to  shield  the 
low  power  stages  from  the  buffer  plate  lines. 
The  under-chassis  shielding  is  completed  when 
a bottom  plate  is  bolted  to  the  chassis. 

The  power  supply  section  occupies  the  tear 
portion  of  the  chassis.  High  voltage  rectifiers 
Di-2-3-4  are  mounted  above  the  deck  (figure 
4)  and  the  diodes  of  the  bias  supply  are 
mounted  below  the  chassis. 

Figure  9 shows  the  interior  of  the  VFO 
compartment.  The  two  osoillator  coils  are 
mounted  on  the  side  of  the  box,  across  from 
the  four  crystal  sockets.  The  trimmer  capacitors 
are  mounted  above  their  respective  coils.  The 
grid  and  cathode  leads  to  oscillator  tube  Vi 
pass  through  a large  rubber  grommet  mounted 
in  the  bottom  of  the  enclosure. 

The  Power  and  The  power  supply  is 
Modulator  Deck  mounted  upon  a steel 
chassis  measuring  13”  x 


6-32  BOLT 

//  N 


A-  3/8*  X 2*  CERAMIC  INSULATOR,  TAPPED  AT  EACH 
END  FOR  6-32  MACHINE  SCREWS. 

B,C-1  1/8*  DIA,  BRASS  DISC  WITH  3/8"  CLEARANCE  HOLE 
FOR  CERAMIC  INSULATOR.  LOWER  DISC  SUPPORTED  ON 
3/8"  CERAMIC  INSULATORS,  USE  FLAT  HEAD  SCREWS. 

D-  3/8"  CERAMIC  INSULATORS, 

E-  1 1/8"  DISC  MADE  OF  TWO  LAYERS  OF  .015"  THICK 
TEFLON  HELD  ATOP  C BY  CELLOPHANE  TAPE  . 

F-  TWO  6/32  NUTS  PLACED  UNDER  STRAP,  STRAP  AND 
NUTS  SOLDERED  TO  BRASS  PLATE  . HOLE  IN  PLATE 
PASSES  6/32  BOLT. 


Figure  13 

CONSTRUCTION  OF 
NEUTRALIZING  CAPACITOR 
FOR  826  STAGE 

Two  capaeiiors  are  required  (Cn  and  Cit), 
Capacitors  hove  top  brace  plate  made  of 
polystyrene  sheet  or  other  good  VHP  insulat- 
ing material.  Plate  is  supported  on  two  cer- 
amic insulators  which  form  center  support  for 
capacitors.  Plates  of  neutralizing  capacitors 
are  made  of  brass  discs.  The  grid  disc  is  sup- 
ported on  short  ceromic  insulators,  and  the 
plate  disc  is  supported  from  the  polystyrene 
plate  by  6-32  machine  screw.  Plate  encircles 
ceromic  insulator  with  a close  fit.  Plate  lead 
and  two  6-32  nuts  are  soldered  to  top  plate. 
Mounting  screw  passes  through  this  assembly, 
and  also  through  hole  drilled  in  top  plate. 
After  capacitors  have  been  adjusted,  top  plate 
is  cemented  to  center  ceramic  insulator  with 
clear  nail  polish  to  retain  ad/ustment. 


17”  X 3”,  as  shown  in  figure  11.  The  modu- 
lator and  rectifier  tubes  are  placed  near  the 
front  panel,  with  the  modulation  transform- 
er and  power  transformer  to  the  rear  of 
the  chassis.  The  swinging  choke  is  placed  be- 
tween the  tubes  at  the  front-center  of  the 
chassis.  Smaller  components  are  mounted  be- 
neath the  chassis.  High  voltages  are  applied  to 
the  transmitter  by  primary  relay  RYi,  con- 
trolled by  switch  Ss  and  an  optional  external 
relay  control  circuit. 

Transmitter  The  tuned  circuits  of  the  trans- 

Adjustment  mitter  can  be  set  to  their  pro- 

per operating  range  with  the 
aid  of  a grid-dip  osoillator.  All  r-f  tubes  must 
be  in  their  sockets.  Coil  La  is  set  to  8.5  Me., 
and  circuit  Ca-Lt  is  adjusted  to  17  Me.  with 
Cz  set  at  half-capacity.  The  7 g/xfd.  trimmer 


HANDBOOK 


De  Luxe  All-Band  Transmitter  673 


capacitor  across  Ls  is  adjusted  to  resonate  the 
plate  circuit  of  the  first  triplet  stage  to  50  Me. 
with  Ca  at  half-capacity,  and  C5-L7  is  tuned  to 
the  same  frequency  in  the  same  manner  with 
50  Me.  plug  PL-1  placed  in  socket  SO-1.  The 
grid  coils  of  the  buffer  stage  are  next  adjusted 
to  frequency. 

The  next  step  is  to  adjust  the  plate  circuit  of 
the  buffer  stage.  Coil  Lio  is  attached  to  the 
transmission  line  about  2"  from  the  feed- 
through terminals,  and  capacitor  C?  is  tapped 
on  the  line  about  1"  closer  to  the  buffer  tube 
than  is  the  coil.  Resonance  of  the  lines  can  be 
adjusted  by  changing  the  length  of  the  flexi- 
ble plate  leads  to  the  buffer  tube,  by  varying 
the  spacing  of  the  lines,  or  by  changing  the 
spacing  between  the  chassis  and  the  lines.  The 
tuning  range  of  C?  may  be  varied  by  shifting 
the  position  of  the  taps  on  the  lines.  Coil  Lw 
is  placed  at  the  "cold”  point  of  the  line,  which 
may  be  found  by  touching  the  line  with  a pen- 
cil lead  when  the  transmitter  is  operating. 
Place  the  coil  at  the  spot  on  the  line  that  has 
the  least  "fire”  when  the  transmitter  is  op- 
erating on  144  Me, 

As  soon  as  grid  current  can  be  measured 
on  the  power  amplifier,  the  stage  may  be  neu- 
tralized. The  transmitter  is  first  neutralized 
on  144  Me.  by  adjusting  capacitors  Cn  and  Cn. 
The  B-plus  lead  should  be  removed  for  this 
operation.  Operating  frequency  is  then  changed 
to  50  Me.  and  link  coils  Lu  and  Lu  are  adjust- 
ed for  complete  neutralization  on  the  six  meter 
band.  The  144  Me.  neutralization  should  then 
be  rechecked.  The  links  have  very  little  ef- 
fect at  144  Me.,  since  the  shorting  bar  effec- 
tively removes  Liz  and  Lu  from  the  circuit. 
However,  the  adjustment  of  the  capacitors  af- 
fects the  setting  of  the  50  Me.  links,  so  the 
latter  must  be  readjusted  after  each  change  in 


the  settings  of  Cn  and  Cn. 

Reduced  voltage  is  now  applied  to  the  am- 
plifier stage  and  it  is  loaded  into  a suitable  an- 
tenna system.  The  links  of  coils  Ls,  Lo.  L?,  and 
Ls  should  be  adjusted  for  maximum  grid  drive. 
Loose  coupling  should  be  used  as  overcoupling 
reduces  excitation  to  the  power  amplifier  and 
results  in  "double  tuning”  of  the  stages.  Buf- 
fer plate  current  runs  about  100  milliamperes, 
and  each  amplifier  tube  should  develop  ap- 
proximately 35  milliamperes  of  grid  current. 

The  144  Me.  antenna  link  (Ln)  is  adjusted 
by  varying  the  length  of  the  loop  and  its  po- 
sition to  the  lines.  It  is  mounted  beneath  the 
lines  and  spaced  about  away  from  them. 
Fine  antenna  loading  may  be  controlled  by 
loading  capacitors  Ca  and  Cio.  The  amplifier 
may  be  loaded  to  200  ma.  for  phone  operation, 
or  250  ma.  for  c.w.  operation.  The  overall  am- 
plifier efficiency  is  better  than  50%,  delivering 
160  watts  to  a dummy  load  with  a power  in- 
put of  300  watts  at  144  Me.  Efficiency  is  bet- 
ter than  60%  at  50  Me. 

Modulator  plate  current  rises  to  about  150 
ma.  on  peaks  for  100%  modulation  of  the  final 
amplifier. 

31-2  A De-Luxe 

Tronsmiffer  for  fhe 
3.5  - 29.7  Me.  Range 

Shown  in  this  section  is  a de-luxe  phone/ 
c.w.  transmitter  for  operation  in  the  80,  40, 
20,  15,  and  10  meter  amateur  bands.  The 
transmitter  runs  500  watts  input  on  c.w.  and 
375  watts  input  on  phone,  and  is  completely 
self-contained  in  two  small  relay  racks  as  il- 
lustrated in  figure  14.  The  r.f.  section  of  the 
transmitter  is  designed  to  be  virtually  TVI- 
proof,  and  modern  techniques  such  as  speech 


Figure  14 

DE-LUXE  STATION 
FOR  THE  DX  MAN 

employes  VFO-controlled, 
bandswiiching,  Phone-CW 
transmitter  running  up  to 
500  watts. 

This  compact,  all-band 
transmitter  employs  a 
7094  tetrode  in  a pi-net- 
work amplifier,  modulat- 
ed by  8il-A*s.  Clickless, 
break-in  keying  is  used 
for  C.W.,  and  speech 
clipping  and  limiting  are 
employed  for  AM  phone. 
Modulator  and  power 
supply  are  in  the  rack 
at  left,  and  entire  r-f 
section  is  in  right-hand 
rack.  High  stability  VFO 
ensures  drift-free  fre- 
quency control.  Transmit- 
ter is  designed  for  max- 
imum harmonic  suppres- 
sion and  is  virtually  TVI- 
proof 


674  Transmitter  Construction 


THE  RADIO 


1 1 5/230  V.  'V 


Figure  15 

BLOCK  DIAGRAM  OF  ALL-BAND  500  WATT  TRANSMITTER 
R-f  section  employs  bandswitching  exciter  for  coverage  of  10,  15,  20,  40,  and  BO  meter  bands.  Variable 
frequency  oscillator  operates  in  1.7  Me.  region  to  achieve  maximum  isolation.  Tuning  and  antenna 
laading  are  aided  by  6ALS  vacuum  tube  voltmeter  monitoring  output  of  ampiifier  stage. 

Low  level  speech  clipping  and  filtering,  combined  with  high  level  audio  filter  imports  "punch"  to  phone 
signal,  yet  holds  sidebands  to  a minimum.  Single  high  voltage  supply  powers  modulator  and  final  amplifier. 


dipping  and  dickless,  break-in  keying  are  in- 
corporated in  the  transmitter. 

A high  stability  variable  frequency  oscilla- 
tor is  employed  for  frequency  control,  the  dial 
of  which  is  directly  calibrated  for  each  band. 
Band  changing  is  a simple  operation  and  may 
be  accomplished  in  a few  seconds. 

Transmitter  Circuitry  A block  diagram  of  the 
and  Layout  transmitter  is  given  in 

figure  15.  Nineteen 
tubes  ate  used  in  the  equipment:  seven  in  the 
r-f  section,  six  in  the  audio  section,  and  six  in 
the  power  supply. 


The  RCA  7094  beam  power  tube  is  em- 
ployed in  the  final  amplifier  stage.  This  com- 
pact mbe  has  high  perveance  and  high  power 
gain.  It  can  be  operated  at  full  input  to  60 
Me.,  and  has  a maximum  plate  dissipation  of 
125  watts.  In  addition,  it  has  triple  base-pin 
connections  for  the  screen  grid  to  permit  good 
r-f  grounding  and  large  plate  radiating  fins 
for  effective  cooling.  The  compact  size  makes 
it  especially  effective  in  the  high  frequency 
portions  of  the  communication  spectrum.  Driv- 
ing requirements  are  modest  and  permit  the 
use  of  a simplified  bandswitching  exciter. 


Figure  16 

VFO-EXCITER  DECK  CONTAINS 
ALL  IMPORTANT  OPERATING 
CONTROLS 

Low  level  /-f  stages  are  built  upon  8"xJ7"x2" 
aluminum  chassis,  Keyer  timing  potentiometer 
Rs  is  at  left,  with  excitation  potentiometer 
Rt,  control  switch  St,  and  bandswitch  Si  in 
line,  left  to  right.  VfO  dial  is  centered  on  the 
panel. 

Output  connector  Pi  is  mounted  on  polystyrene 
plate  atop  shielded  enclosure  at  right.  Shield- 
ing is  made  from  Reynold's  "Do-it-yourself" 
perforated  aluminum  sheet. 


V 2 Va  V4 


Cl — 300  p-iifd.  Bud  Cf-2008  dual  bearing  capacitor 

Cs — ^300  jx/j-fd.  silver  mica  capacitor.  Vary  for  desired  frequency  coverage. 

C3 — 2000  fx/^fd.  High  accuracy  capacitor,  tolerance  1%,  temperature  coefficient 
zero,  plus  or  minus  10  p.p.m.  Centralab  950-202 
Ci,  Cs,  Ce — 25  /x/xfd.  midget  variable  padding  capacitor,  ceramic 
Li — 79  turns  #20  e.  wire,  tapped  5 turns  from  ground  end.  Wound  on  ceramic 
form,  diam.,  V*  high.  Space  the  winding  by  interwinding  #28  e.  wire, 
thert  remove.  National  XR-62  form. 

Li — 30  turns  it28  dsc.,  Y2"  diam.,  Yz*'  long  on  National  XR-50  form 


Lj — 21  turns  #22  e.,  Y2'*  diam.,  Yi*^  long  on  National  XR-SO  form 
5iA-B-C^D — Two  pole,  6 position  decks  (Centralab  PA-3)  with  Centralab  PA-301 
index  assembly. 

5tA-B-C — 3 pole,  3 position  rotary  switch 
RYt — DPDT  relay,  7,000  ohm  d.c.  coil 
RY} — DPDT  relay,  1 15-volt  a.c.  coil 

PC — 52  ohm,  1-watt  carbon  resistor,  wound  with  3 turns  ^20  e.  wire 
RFCi — 2.5  mh.  choke  National  R-100 
RFCi — VHF  choke.  Ohmite  Z-144 


z 

> 

z 

D 

00 

O 

o 


De  Luxe  All-Band  Transmitter  675 


Figure  18 

INTERIOR  OF  VFO 
SHOWING 
PLACEMENT  OF 
MAJOR 
COMPONENTS 

The  VFO  is  built  upon 
two  Ya’inch  durai  piates, 
one  forming  the  base 
and  the  other  forming 
the  front  of  the  osciiiator 
enclosure,  inductor  Li 
and  padding  capacitors 
are  mounted  to  front 
wait,  and  the  variabte 
tuning  capacitor  is  boit~ 
ed  to  base.  Osciiiator 
tube  (V-1)  projects  from 
far  wait  of  enclosure  to 
remove  as  much  heat 
from  tuned  circuit  as 
possibte.  Entire  VFO  is 
mounted  upon  miniature 
shock  mounts  and  dial  is 
attached  to  VFO  unit, 
rather  than  to  rack 
panet.  6AG7  and  2E26 
tubes  are  at  teft,  with 
2E26  ptate  choke  and 
5KV  coupling  capacitor 
in  foreground. 


The  variable  frequency  oscillator  and  exci- 
ter stages  are  built  upon  one  chassis  deck,  as 
shown  in  figures  16,  18,  19,  and  20.  The 
complete  schematic  of  this  unit  is  given  in 
figure  17.  The  variable  oscillator  covers  the 
range  of  1.75  Me. -1.85  Me.,  permitting  full 
coverage  of  all  amateur  bands  except  the  ex- 
treme top  of  the  28  Me.  band,  and  the  top 
portion  of  the  80  meter  band.  Readjustment 
of  the  oscillator  padding  capacitor  provides 
complete  coverage  of  these  segments,  if  de- 
sired. The  VFO  consists  of  a 6U8  pentode- 
triode  (Vi)  operating  as  a "hot  cathode"  oscil- 
lator and  triode  cathode  follower.  Extremely 
high-C  is  employed  in  the  tuned  circuit  of  the 
VFO  to  obtain  a good  order  of  stability.  A 
2000  g/ifd.  precision  ceramic  capacitor  G, 
forms  the  large  portion  of  the  tuning  capaci- 
tance, and  the  coil  Li  is  wound  upon  a ceramic 
form  having  a very  low  temperature  coeffi- 
cient. The  cathode  follower  stage  provides 
excellent  circuit  isolation  for  the  oscillator, 
combined  with  a minimum  of  plate  circuit 
loading.  Figure  18  shows  the  layout  of  the  ma- 
jor oscillator  components.  The  circuit  is  built 
upon  an  Vs”  thick  aluminum  plate,  bolted  to 
a small  panel  of  the  same  material.  The  sides, 
back,  and  top  of  the  oscillator  enclosure  are 
made  of  aluminum  sheet.  The  complete  unit 
is  mounted  upon  miniature  rubber  shock 
mounts.  The  VFO  dial  is  firmly  affixed  to  the 
enclosure  and  actually  does  not  make  physical 
contact  with  the  panel  at  all.  The  frequency 
determining  unit  is  thereby  protected  from  jolts 
and  jars  which  might  cause  a "warble”  in  the 
transmitter  frequency.  The  r-f  output  from  the 
cathode  follower  is  5 volts,  r.m.s. 


Three  tubes  are  employed  in  the  multiplier 
and  driver  stages  of  the  transmitter.  The  first 
6AC7  doubler  stage  (Vz)  operates  into  a 
broadly  resonant  slug-mned  coil  Lz.  When 
tuning  the  exciter  unit,  this  coil  is  peaked  at 
3.6  Me.  The  stage  will  then  deliver  substan- 
tially constant  output  over  the  range  from  3.2 
to  4.0  Me.  The  2E26  stage  (Vi)  operates  as 
an  amplifier  on  the  3.5  Me.  band  and  as  a 
doubler  on  the  7 Me.  band,  driven  directly 
from  the  6AC7  buffer-doubler. 


Figure  19 

REAR  VIEW  OF  EXCITER  DECK 

MuHip/rer  tubes  are  in  the  foreground.  Ad- 
justment holes  for  the  multiplier  trimmer  ca- 
pacitors are  on  side  of  chassis.  1/2 -inch  angle 
stock  is  run  around  chassis  and  ponef  to  form 
foundation  to  which  the  perforated  enclosure 
is  attached.  Keyer  tube  V?  is  placed  in  rear 
corner.  Potentiometer  Rg  and  auxiliary  key 
jack  are  on  chassis  wall  below  keyer  tube. 
Short  lead  extends  from  2f26  blocking  ca- 
pacitor up  to  connecting  plug  Pi  mounted  on 
enclosure. 


De  Luxe  All-Band  Transmitter  677 


Figure  20 

UNDER-CHASSIS  VIEW  OF  EXCITER 


All  power  wiring  is  run  in  shielded  braid,  grounded  to 
the  chassis  ai  convenient  points.  Voltage  regulator 
tube  Vik  is  mounted  beneath  the  chassis  on  smalt 
angle  brackets.  RYi  and  RYs  are  mounted  to  the  inner 
chassis  wall,  directly  behind  keyer  panel  control  Rj. 
R-f  circuits  are  at  the  opposite  end  of  chassis,  groups 
around  bandswitch  Su  The  slugs  of  coils  Lj  and  Lj  are 
adjustable  from  atop  the  chassis.  Metal  plate  covers 
bottom  of  chassis  to  reduce  spurious  radiation  from 
r.f.  circuits. 


For  operation  on  20,  15,  and  10  meters, 
switch  SiD  disconneas  the  2E26  from  the 
6AC7  stage,  and  connects  it  to  an  auxiliary 
6AG7  frequency  multiplier  (Va).  The  multi- 
plier operates  as  either  a doubler,  a tripler,  or 
as  a quadruplet.  Its  output  circuit,  therefore,  is 
either  tuned  to  7 Me.,  10.5  Me.,  or  14  Me. 
Coil  Ls  in  the  plate  circuit  of  the  6AG7  re- 
sonates with  residual  circuit  capacities  to  14 
Me.  Switch  section  SiC  of  the  bandswitch  adds 
additional  capacity  to  lower  the  resonant  fre- 
quency of  this  circuit  to  10.5  Me.  or  7 Me. 

The  2E26  stage  operates  as  a doubler  to  14 
Me.,  receiving  7 Me.  excitation  from  the  6AG7 
stage.  When  the  6AG7  is  tuned  to  10.5  Me., 
the  2E26  serves  as  a doubler  to  the  21  Me. 
band.  For  28  Me.  output  from  the  2E26,  14 
Me.  excitation  is  received  from  the  6AG7 
stage,  and  the  2E26  again  functions  as  a 
doubler. 

Keying  and  A simplified  schematic  of  the 

Control  Circuit  keying  and  control  circuit 

for  the  exciter  is  shown  in 
figure  21.  A break-in  type  keyer  is  employed 
for  c.w.  operation.  The  unit  is  completely  actu- 
ated by  the  transmitter  key.  When  the  key  is 
pressed  the  transmitter  is  energized  and  may 
be  keyed  in  a normal  manner.  The  keyer 
"holds”  the  transmitter  power  supplies  on 
while  the  key  is  manipulated.  When  the  op- 
erator stops,  the  keyer  turns  off  the  transmitter 
after  a short  pause.  Thus,  the  transmitter  re- 
mains on  between  keyed  characters  and  words. 


KEYER  AND  CONTROL  CIRCUITS 

Clickless,  break-in  keying  is  provided  by  this  simple  control  circuit.  See  text  for  full  details  of  operation 


678  Transmitter  Construction 


THE  RADIO 


Figure  22 

500  WATT  POWER 
AMPLIFIER  HAS 
SYMMETRICAL 
PANEL  LAYOUT 

The  7094  amplifier  stage 
Is  enclosed  in  a separate 
deck.  Plate  milliammeter 
is  at  left,  with  0-1  d.c. 
milliammeter  at  right 
which  may  be  inserted  in 
various  power  leads  in 
exciter  or  amplifier. 
Large  center  knobs  are 
(left)  pi-network  switch 
and  (right)  main  ampli- 
fier tuning  control. 
Controls  across  the  bot- 
tom are  (left  to  right): 
Pi-network  loading  ca- 
pacitor, auxiliary  pi-net- 
work capacitor  switch, 
grid  circuit  tuning  ca- 
pacitor, and  meter 
switch.  Pi-network  switch 
Si  is  ganged  with  grid 
turret  switch  Ss  elimin- 
ating one  panel  control. 


but  will  return  to  a stand-by  position  after  the 
keying  action  has  been  completed.  The  "turn- 
off” time  may  be  varied  to  suit  the  operator’s 
taste.  For  phone  operation,  the  time  delay  cir- 
cuit is  eliminated,  and  the  push-to-talk  circuit 
actuates  the  keyer  and  control  relays  directly. 
The  whole  sequence  may  be  over-ridden  and 
manual  operation  restored  by  means  of  a "man- 
ual-automatic” switch  on  the  power  supply 
chassis.  Finally,  a "transmit  - tune  - standby” 
switch  S2A  removes  screen  voltage  from  the 
final  amplifier  stage  for  tune-up  purposes. 

Referring  to  figure  21,  the  sequence  of  op- 
eration is  as  follows: 

1 —  Key  up,  switch  S2B  on  "transmit.” 

The  right-hand  section  of  the  keyer  tube 
V?  is  cut  off  by  the  negative  bias  on  the 
cathode.  Potentiometer  R2  is  adjusted  for 
1 ma.  current  through  15K  resistor  to 
ground  (-15  volts  to  chassis  at  point  A). 
6AC7  (V2)  is  cut  off. 

2 —  ^Key  is  depressed. 

Right-hand  section  of  V?  conducts  heavily 
and  cathode  voltage  rises  quickly,  cutting 
off  left  section  of  V?  and  removing  block- 
ing bias  of  6AC7  (V2).  Plate  current  of 
right-hand  section  of  V7  operates  RY2  and 
charges  40  gfd.  capacitor  through  1N92 
diode.  The  relay  closes,  placing  B-plus 
voltage  on  oscillator  stage  and  operating 


the  high  voltage  primary  circuit  relay  cir- 
cuit. This  circuit  may  be  rendered  inoper- 
ative on  "standby”  position  by  switch  S2B, 
enabling  VFO  to  be  used  for  frequency 
check  or  "spotting”  purposes. 

3 —  Key  is  released  . 

Right-hand  seaion  of  V7  cuts  off  immed- 
iately. Left-hand  section  draws  1 ma.  and 
cuts  off  6AC7  buffer  tube.  The  1N92  is 
back-biased  by  rise  in  12AU7  plate  volt- 
age, keeping  charge  on  40  gfd.  capacitor. 
Relay  RY2  remains  closed  until  capacitor 
is  discharged  through  shunting  resistances. 
Relay  then  opens,  turning  off  high  voltage 
supplies  of  transmitter.  "Release”  delay  is 
controlled  by  potentiometer  Ra. 

4 —  Phone  Operation 

Switch  Si  on  modulator  deck  is  closed  for 
phone  operation,  placing  the  coil  of  relay 
RYa  in  the  circuit.  When  the  key  is  closed, 
or  the  press-to-talk  circuit  activated,  RYaA 
shorts  out  the  time  delay  circuit  of  RY2, 
and  at  the  same  time  RYsB  removes  a 
short  across  the  screen  modulation  choke 
in  the  final  amplifier. 

Exciter  plate  voltage  is  obtained  from  the 
modulator  power  supply  (See  Chapter  30, 
figure  17),  and  bias  and  amplifier  voltage  are 
derived  from  the  main  power  supply,  located 
on  a separate  chassis.  The  plate  relay  RYi  may 
be  actuated  by  switch  Sa  or  by  the  keyer  circuit. 


HANDBOOK 


De  Luxe  All-Band  Transmitter  679 


V5  00 1 

7094  5KV  Ls  J2 


C» — Disc-type  neutralizing  capacitor.  National  NC-800  or  equivalent. 

Ciu — 750  piiftl.,  3000  volt.  Johnson  1S0C30 

Cn — 900  piild.  Two  gang  broadcast  capacitor  (see  text). 

Cii — Three  .001,  5KV  ceramic  capacitors.  Centralab  type  850.  Place  one  capacitor  on  each  screen  terminal. 
Li — Composed  of  two  coils  mounted  on  back  of  Si  and  connected  in  series. 

80  meter  coil;  28  turns,  I"  diam.,  Ifi"  long.  40  meter  tap  at  18th  turn  from  "cold"  end. 

20  meter  coil;  7 turns,  1"  diam.,  V/i"  long.  15  meter  tap  at  4 turns,  10  meter  tap  at  3 turns  from 
"cold"  end.  Adjust  taps  with  grid-dip  oscillator. 

Ls-5i — Barker  & Williamson  #3850,  500  watt  turret  tor  10-80  meters. 

5j — Single  pole,  6 position  ceramic  switch,  progressively  shorting.  Centralab  PA-2042. 

RFCt — 2.5  mh.  choke.  National  R-100 

RFC] — Heavy  duty  choke  for  3-30  Me.  range,  500  mo.  Ray  par  R1.-702 

RFCi — VHF  Choke,  Ohmite  Z-144,  or  equivalent 

PC — 50  ohm  composition  resistor  wound  with  3 turns  #20  e. 

T I — 6.3  volts  at  4 a.  Stancor  P-4019 

CHt — 7 h.  at  50  mo.  Stancor  C-170T 

RYi — SP5T  sensitive  relay,  with  1 milliampere  d.c.  coil 

Ml — 0 - 500  d.c.  milliammeter  with  insulated  zero-set  and  bakelite  case. 

B — Small  1 IS-volt  a.c.  blower  motor. 


The  Amplifier  The  schematic  of  the  ampli- 
Sroge  fier  is  shown  in  figure  23,  and 

various  layout  photographs  are 
given  in  figures  22,  24,  25,  and  26.  Par- 
ticular attention  is  given  to  TVI-reduction 
measures  in  the  construction  of  the  unit.  A 
single  7094  beam  power  tube  is  used  in  a pi- 
network  circuit.  The  grid  circuit  (L4-G)  is 
common  to  the  plate  circuit  of  the  2E26  and 
is  capable  of  complete  frequency  coverage  by 
virtue  of  the  tapped  coll  assembly  shown  in 
the  under-chassis  view.  Bridge  neutralization 
is  employed  to  achieve  maximum  stability. 
Variable  capacitor  C9  and  the  grid  bypass  ca- 
pacitor Cs  form  two  legs  of  the  bridge. 

To  take  advantage  of  the  triple  base-pin 
connections  to  the  screen  of  the  7094  three 
separate  low  inductance  bypass  capacitors  (C12) 
are  used  in  the  screen  circuit,  one  attached 


to  each  of  the  screen  socket  terminals.  A VHF 
parasitic  choke  (PC)  is  included  In  the  lead 
to  the  control  grid  of  the  tube.  As  a result  of 
these  precautions,  the  amplifier  is  completely 
stable  over  the  whole  operating  range. 

A sensitive  relay  RYs  Is  Incorporated  in  the 
screen  circuit  to  protect  the  tube  under  condi- 
tions of  low  plate  potential.  The  relay  will 
open  the  screen  circuit  when  the  plate  poten- 
tial Is  removed  from  the  amplifier  stage.  By 
far  the  most  frequent  type  of  damage  to  a 
tetrode  tube  is  caused  when  full  screen  volt- 
age is  applied  in  the  absence  of  plate  voltage. 
The  screen  relay  protects  the  tube  from  this 
type  of  damage.  Self-modulation  of  the  screen 
is  accomplished  by  an  iron  core  choke  CH,  in 
the  screen  circuit.  This  choke  is  shorted  out 
by  relay  RYaB  (figure  21 ) for  c.w.  operation. 

A pi-network  configuration  is  used  in  the 


680  Transmitter  Construction 


THE  RADIO 


Figgre  24 

REAR  OBLIQUE  VIEW  OF 
PI-COUPLED  AMPLIFIER 

Dural  angle  strips  are  mounted  around  edge 
of  8'^x17"x2'*  aluminum  chassis  and  panel  to 
which  TVI  enclosure  is  attached  with  self- 
tapping  sheet  metal  screws.  Meters  are  en- 
cased in  aluminum  cups,  with  r.f.  filters 
mounted  on  the  back  of  each  case.  Meter 
leads  are  10  KV  television  cable,  enclosed  in 
shielded  braid.  The  7094  tetrode  is  mounted 
at  rear  of  chassis,  with  neutraliiing  capacitor 
Cs  to  left,  and  plate  choke  at  right.  Connec- 
tions to  plate  of  tube  are  made  with  flexible 
copper  strap. 

Filament  transformer  is  at  left  of  chassis,  with 
blower  motor  hidden  behind  it.  Output  capaci- 
tor of  pi-network  is  below  deck. 


plate  circuit  of  the  amplifier,  employing  one 
of  the  new  Barker  & Williamson  500  watt 
band-switching  assemblies.  A two-gang  broad- 
cast capacitor  is  employed  for  the  low  im- 
pedance output  loading  control.  Switch  Ss  al- 
lows the  insertion  of  an  extra  1000  /x^fd.  ca- 
pacitor for  operation  into  low  impedance  loads 
at  3.5  Me.  The  output  circuit  of  the  network 
is  monitored  by  a 6AL5  vacuum-tube  volt- 
meter (Vo)  which  may  be  employed  during 
tune-up  and  loading  adjustments. 

All  power  leads  beneath  the  chassis  are 
shielded  in  flexible  braid,  grounded  at  con- 
venient points  to  the  chassis.  The  meters  are 
enclosed  in  spun  aluminum  cups  that  are  bolt- 
ed to  the  panel  of  the  stage.  Each  meter  lead 
is  bypassed  to  the  chassis  with  a .001,  5KV 
ceramic  capacitor  and  a VHP  tjioke  is  placed 
in  series  with  the  meter  terminal.  A small  115- 
volt  blower  motor  is  mounted  at  one  end  of 
the  chassis  (figure  25)  to  direct  a cooling 
movement  of  air  across  the  envelope  of  the 
tube. 

Placement  of  the  major  components  beneath 
the  chassis  may  be  seen  in  figure  26.  An 
aluminum  partition  separates  the  pi-network 
components  from  the  tube  socket  and  the  grid 
circuit.  The  grid  coil  switch  (Ss)  is  driven  by 


means  of  a dial  cord  and  drum  from  the  plate 
bandswitch  inductor  shaft  (Ls)  atop  the  chas- 
sis. The  cord  passes  through  two  holes  in  the 
chassis  deck.  The  main  bandswitch  control  on 
the  amplifier  panel  thus  actuates  both  the 
grid  and  plate  switches.  Safety  relay  RYs  is 
mounted  on  the  tear  apron  of  the  chassis,  and 
an  aluminum  plate  covers  the  complete  under- 
chassis  area.  A one-inch  hole  is  cut  in  this 
plate  directly  below  the  stator  terminal  of  grid 
mning  capacitor  Q,  and  a banana  jack  is 
mounted  to  a polystyrene  plate  bolted  in 
place  over  the  hole.  This  forms  the  low  capaci- 
ty input  terminal,  Ji.  A mating  hole  is  cut  in 
the  top  of  the  exciter  screen  and  covered  with 
a second  polystyrene  plate  having  a banana 
plug  mounted  in  the  center.  When  the  ampli- 
fier is  slipped  in  the  rack  the  plug  and  jack 
mate,  forming  the  connection  between  the 
2E26  plate  and  the  tuned  tank  in  the  grid 
circuit  of  the  final  amplifier. 

Power  Supply  The  power  supply  is  built 
and  Modulator  upon  a separate  chassis,  as 
shown  in  figure  27.  Two 
866A  reaifier  tubes  are  employed  with  a 
choke  input  filter  to  supply  1500  volts  at 
500  milliamperes  for  the  r-f  amplifier  and 
modulator.  "Hash”  suppression  chokes  are 
placed  in  the  rectifier  plate  leads  to  reduce  the 
high  frequency  "buzz”  sometimes  super-im- 
posed upon  the  carrier  by  mercury  vapor  tubes. 
Sufficient  capacity  must  be  used  in  the  filter 
system  to  afford  good  dynamic  stability  to  the 
supply  under  keyed  loads,  or  under  conditions 


Figure  25 

LEFT  OBLIQUE  VIEW  OF 
PI-COUPLED  AMPLIFIER 

Blower  motor  to  cool  7094  tube  is  located 
between  front  panel  and  filament  transform- 
er. High  voltage  and  coaxial  antenna  terminals 
are  on  right  of  rear  wall  of  chassis.  Addition 
of  perforated  screen  makes  enclosure  radia- 
tion-proof. 


HANDBOOK 


of  heavy  modulation.  Eight  60  /xfd.,  450  volt 
tubular  electrolytic  capacitors  are  connected  in 
series  parallel  to  form  a 30  /xfd.,  1800  volt 
unit.  A small  bias  supply  delivering  -105 
volts  for  the  7094  amplifier  and  keyer  is  also 
included  on  the  chassis.  Layout  and  wiring  of 
the  supplies  are  shown  in  figures  27  and  29. 

The  modulator  has  been  previously  shown 
in  Chapter  30,  figure  17.  The  speech  amplifier 
supply  provides  370  volts  for  exciter  operation. 
One  section  of  the  "phone-c.w.”  relay  (RYaA) 
shorts  out  the  time  delay  circuit  on  the  exci- 
ter deck  for  phone  operation,  and  a high  volt- 
age relay  RYi  is  employed  to  remove  the  mod- 
ulator for  c.w.  operation.  Six  volts  is  also  sup- 
plied to  the  exciter  for  filament  service.  This 
filament  circuit  is  balanced  to  ground  to  re- 
duce residual  hum  in  the  speech  amplifier,  so 
the  ground  circuit  cannot  be  employed  as  a 
filament  return  in  the  exciter. 

Exciter  Tune-up  Each  deck  of  the  transmitter 
and  Operation  should  be  tested  as  an  indi- 
vidual unit  before  the  trans- 
mitter is  tested  in  entirety.  Screen  voltage  of 
the  2E26  stage  should  always  be  set  at  xero 
by  means  of  potentiometer  Ri  (figure  17) 
when  the  exciter  is  operated  without  the  final 
amplifier,  since  the  2E26  plate  circuit  is  in- 
complete, and  excessive  screen  current  will  be 
drawn  by  the  2E26  unless  the,  screen  potential 
is  removed. 

Oscillator  calibration  may  be  adjusted  by 
varying  the  spacing  of  the  turns  of  coil  Li,  or 


Figure  27 

REAR  VIEW  OF  POWER  SUPPLY 
Primary  power  leads  pass  through  bulkhead- 

iype  filter  capacitors  to  prevent  radiation  of 
TVI-producing  harmonics  from  power  line. 
Voltage  regulator  tube  for  bios  supply  is 
visible  in  corner  of  chassis. 


Figure  26 

UNDER-CHASSIS  VIEW  OF 
POWER  AMPLIFIER 

All  power  wiring  beneath  the  chassis  is  placed 
within  braid  to  reduce  pickup  of  r.f.  energy. 
Note  that  shield  divides  area  into  two  sections. 
Smaller  area  contains  output  capacitor  of  pi~ 
network,  6ALS  diode  voltmeter  and  high  volt- 
age leads.  Larger  area  contains  grid  circuit 
components  and  meter  switch.  Sensitive  relay 
RYs  is  mounted  to  wall  of  the  chassis  behind 
the  grid  tuning  capacitor.  Three  .001,  5 Kv. 
ceramic  screen  bypass  capacitors  are  mounted 
close  to  socket  terminals.  Grid  turret  switch  is 
driven  from  plate  turret  atop  chassis. 


by  changing  the  value  of  the  auxiliary  padding 
capacitor  Cz.  Alignment  of  the  6AC7  and 
6 AG  7 stages  is  done  by  observation  of  the 
grid  current  of  the  2E26  stage,  as  measured 
across  shunt  resistor  Ri  in  the  grid  return  cir- 
cuit. This  circuit  is  not  connected  to  the  meter 
switch,  since  it  requires  only  one  preliminary 
reading  during  initial  tune-up.  Coil  Lz  is 
peaked  at  about  3.6  Me.  with  padding  capaci- 
tor about  half  meshed.  If  grid  current  to  the 
2E26  is  excessive,  the  15K  resistor  shunted 
across  coil  Lz  may  be  reduced  in  value.  Grid 
current  should  remain  between  2 and  3 mil- 
liamperes  across  any  band. 

After  80  meter  alignment  is  completed, 
bandswitch  Si  is  set  to  the  10  meter  position. 
The  slug  of  coil  Ls  is  adjusted  for  maximum 
grid  current  across  the  28  Me.  band.  Coil  Lz 
resonates  to  14  Me.  for  this  operation.  The 
bandswitch  is  next  set  to  the  1 5 meter  position 
and  the  VFO  tuned  to  21.2  Me.  Padding  ca- 


682  Transmitter  Construction 


THE  RADIO 


866-A  CHl 


SCHEMATIC,  TRANSMITTER  POWER  SUPPLY 

Cl,  C, — Bulkhead  type,  0.1  iitd.  600  volt  capacitor.  Sprague  "Hypass,"  or  eguiyatent 

CHl — 10  henry  at  300  ma.  Chicago  R.'IOS 

RFC — "Hash"  suppression  choke.  Millen  77866 

RYi — DPDT  relay,  20  amp.  contacts,  115  y.  a-c  coil 

SOi; — eight  pin  receptacle,  CInch-Jones 

Ti— 1710/1430  each  side  c.i.,  300  mo.,  CCS.  Chicago  P-1512 
Ti — 2.5  V.,  10  a.,  10  KV  insulation.  Stancor  P-3060 
Ti — 125  y.,  SO  mo.,  6.3  v.,  2o.  Sfoneor  PA-8421 


pacitor  Cs  is  adjusted  for  correct  grid  current 
across  the  21  Me.  band.  As  a last  step,  the 
bandswitch  is  set  to  20  meters,  the  VFO  tuned 
to  14.2  Me.  and  capacitor  Cs  adjusted  for  uni- 
form grid  drive  across  the  20  meter  band. 

The  consistency  of  grid  drive  across  the 
various  amateur  bands  is  controlled  by  the 
shunt  resistors  placed  across  coils  Li  and  Li. 
The  lower  the  value  of  these  resistors,  the  more 
uniform  is  the  grid  drive  across  the  various 
bands.  At  the  same  time,  the  grid  current  of 
the  2E26  stage  drops  as  these  resistors  are 
lowered  in  value.  The  values  given  in  the  sche- 
matic give  good  results  on  all  bands,  since 
the  grid  current  of  the  2E26  may  vary  over 
a 2 : 1 ratio  across  an  amateur  band  with  little 
change  in  output  from  the  stage. 

Keying  of  the  exciter  is  smooth  and  click- 
less. Buffer  keying  is  controlled  by  the  ad- 
justment of  potentiometer  Rz.  The  buffer 
should  turn  on  before  the  2E26'  stage  ,and  re- 
main on  after  the  2E26  has  been  cut-off.  Time 
delay  before  the  transmitter  is  turned  off  after 
keying  is  controlled  by  potentiometer  Ra.  Cur- 
rent flowing  through  relay  coil  RYz  may  be 
varied  by  changing  the  two  plate  load  resis- 


tors of  the  right-hand  section  of  keyer  tube 
Vi.  The  plate  load  resistor  is  made  up  of  a 
15K  and  a lOK  resistor.  The  lOK  resistor  may 
be  removed  and  a lOK  relay  coil  may  be  in- 
serted in  its  place  if  an  extra  relay  is  desired 
for  operation  of  auxiliary  circuits.  Spring  ten- 
sion of  relay  RYz  may  be  varied  for  optimum 
relay  operation  after  the  circuit  has  been  ad- 
justed. 

Exciter  keying  should  be  relatively  soft  as 
the  addition  of  the  amplifier  stage  will  sharp- 
en it  to  some  extent.  The  0.1  gfd.  waveform- 
ing capacitor  in  the  bias  return  lead  of  the 
2E26  may  be  increased  in  value  to  soften  the 
exciter  keying  to  compensate  for  the  sharpen- 
ing action  of  the  class-C  amplifier  following 
the  keyed  stage.  In  actual  practice,  the  keying 
of  the  exciter  should  be  adjusted  so  that  dots 
run  together  when  sent  on  a bug  at  a rate  of 
about  25  w.p.m. 

The  keyer  may  be  disabled  if  it  is  desired 
to  make  adjustments  to  the  r.f.  section  of  the 
exciter.  Keyer  tube  V,  is  removed  from  the 
socket  and  pins  1 and  7 are  grounded.  Relay 
RYz  is  then  held  closed  by  placing  a piece  of 
cardboard  under  the  back  contacts. 


HANDBOOK 


De  Luxe  All-Band  Transmitter  683 


Figure  29 

UNDER-CHASSIS  LAYOUT  OF 
PARTS  IN  POWER  SUPPLY 

The  866-A  sockets  are  sub-mounted  on 
inch  ceramic  insulators.  Eight  electrolytic  ca- 
pacitors are  mounted  on  phenolic  plate  and 
banded  with  insulated  strap  to  form  high 
voltage  filter  capacitor.  Bias  transformer  is 
next  to  rectifier  tube  sockets.  All  high  voltage 
connections  are  made  with  5 KY  television- 
type  wire. 


Preliminary  The  amplifier  stage  may  be 

Amplifier  neutralized  by  supplying  exci- 

Adjustment  ration  to  the  input  circuit  until 
rectified  grid  current  flows.  A 
jumper  is  placed  across  the  antenna  output 
jack  Ji  and  plate  tuning  capacitor  Go  is  reson- 
ated to  frequency.  Neutralizing  capacitor  O 
is  now  adjusted  until  the  tuning  of  the  plate 
capacitor  has  no  effert  upon  the  reading  of 
the  rectified  grid  current. 

Preliminary  tune-up  into  an  antenna  may 
be  done  with  the  control  switch  (S;,  figure  17) 
in  the  "stand-by”  position.  Screen  voltage  is 
removed  from  the  final  amplifier  screen  circuit. 
Also,  screen  voltage  cannot  be  applied  to  the 
tube  unless  relay  RYs  (fig.  23)  is  closed  by 
applying  plate  voltage  to  the  amplifier  stage. 


PL7  PL4 

R.F.  AMPLIFIER  MODULATOR 


GROUND  LEAD. 


Figure  30 

INTERCONNECTING 
CABLES  FOR 
TRANSMITTER 


Control  Cables  After  testing  is  completed, 
and  Power  Leads  the  units  should  be  placed 
in  the  racks  or  cabinets  and 
control  cables  made  up  as  shown  in  figure  30. 
Wire  sizes  should  be  observed,  as  several  of  the 
leads  carry  large  values  of  filament  current.  In 
addition,  a separate  grounding  lead  should  be 
run  between  the  individual  chassis  as  a safety 
measure.  Do  not  depend  upon  a metal  relay 
rack  for  a ground  return.  All  external  cables 
should  be  laced  for  neat  appearance.  A typical 
layout  of  this  equipment  is  shown  in  figure  14. 


CHAPTER  THIRTY-TWO 


In  view  of  the  high  cost  of  iron-core  com- 
ponents such  as  go  to  make  up  the  bulk  of  a 
power  supply,  it  is  well  to  consider  carefully 
the  design  of  a new  or  rebuilt  transmitter  in 
terms  of  the  minimum  power  supply  require- 
ments which  will  permit  the  desired  perform- 
ance to  be  obtained  from  the  transmitter.  Care- 
ful evaluation  of  the  power  supply  require- 
ments of  alternative  transmitter  arrangements 
will  permit  the  selection  of  that  transmitter 
arrangement  which  requires  the  minimum  of 
power  supply  components,  and  which  makes 
most  efficient  use  of  such  power  supplies  as 
are  required. 

32-1  Power  Supply 
Requirements 

A power  supply  for  a transmitter  or  for  a 
unit  of  station  equipment  should  be  designed 
in  such  a manner  that  it  is  capable  of  deliver- 
ing the  required  current  at  a specified  voltage, 
that  it  has  a degree  of  regulation  consistent 
with  the  requirements  of  the  application,  that 
its  ripple  level  at  full  current  is  sufficiently 
low  for  the  load  which  will  be  fed,  that  its  in- 
ternal impedance  is  sufficiently  low  for  the 


job,  and  that  none  of  the  components  shall  be 
overloaded  with  the  type  of  operation  contem- 
plated. 

The  meeting  of  all  the  requirements  of  the 
previous  paragraph  is  not  always  a straight- 
forward and  simple  problem.  In  many  cases 
compromises  will  be  involved,  particularly 
when  the  power  supply  is  for  an  amateur  sta- 
tion and  a number  of  components  already  on 
hand  must  be  fitted  into  the  plan.  As  much 
thought  and  planning  should  be  devoted  to  the 
power-supply  complement  of  an  amateur  sta- 
tion as  usually  is  allocated  to  the  r-f  and  a-f 
components  of  the  station. 

The  arrival  at  the  design  for  the  power  sup- 
ply for  use  in  a particular  application  may  best 
be  accomplished  through  the  use  of  a series 
of  steps,  with  reference  to  the  data  in  this 
chapter  by  determining  the  values  of  compo- 
nents to  be  used.  The  first  step  is  to  estab- 
lish the  operating  requirements  of  the  power 
supply.  In  general  these  are: 

1.  Output  voltage  required  under  full  load. 

2.  Minimum,  normal,  and  peak  output  cur- 
rent. 

3.  Voltage  regulation  required  over  the  cur- 
rent range. 


684 


Requirements  685 


Figure  1 

POWER  SUPPLY 
CONTROL  PANEL 

A well  designed  supply  con> 
irol  panel  has  separate  primary 
switches  and  indicator  lamps 
for  the  filament  and  plate  cir- 
cuits, overload  circuit  breaker, 
plate  voltage  control  switch 
and  primary  circuit  fuses. 


PLATE  VOLTAGE 


FILAMENT 

VOLTAGE 


PLATE 

VOLTAGE 


O LINE  VOLTAGE  o 


OVERLOAD  EMISSION 

RFIAY  FINAL  BIAS  SCREEN  MOO  FONECW 


4.  Ripple  voltage  limit. 

5.  Rectifier  circuit  to  be  used. 

The  output  voltage  required  of  the  power 
supply  is  mote  or  less  established  by  the  oper- 
ating conditions  of  the  tubes  which  it  will  sup- 
ply. The  current  rating  of  the  supply,  however, 
is  not  necessarily  tied  down  by  a particular 
tube  combination.  It  is  always  best  to  design 
a power  supply  in  such  a manner  that  it  will 
have  the  greatest  degree  of  flexibility;  this 
procedure  will  in  many  cases  allow  an  exist- 
ing power  supply  to  be  used  without  change 
as  a portion  of  a new  transmitter  or  other  item 
of  station  equipment.  So  the  current  rating  of 
a new  power  supply  should  be  established  by 
taking  into  consideration  not  only  the  require- 
ments of  the  tubes  which  it  immediately  will 
feed,  but  also  with  full  consideration  of  the 
best  matching  of  power  supply  components  in 
the  most  economical  current  range  which  still 
will  meet  the  requirements.  It  is  often  long- 
run  economy,  however,  to  allow  for  any  likely 
additional  equipment  to  be  added  in  the  near 
future. 

Current-Rating  The  minimum  current  drain 
Considerations  which  will  be  taken  from  a 
power  supply  will  be,  in  most 
cases,  merely  the  bleeder  current.  There  are 
many  cases  where  a particular  power  supply 
will  always  be  used  with  a moderate  or  heavy 
load  upon  it,  but  when  the  supply  is  a portion 
of  a transmitter  it  is  best  to  consider  the  mini- 


mum drain  as  that  of  the  bleeder.  The  mini- 
mum current  drain  from  a power  supply  is  of 
importance  since  it,  in  conjunction  with  the 
nominal  voltage  of  the  supply,  determines  the 
minimum  value  of  inductance  which  the  input 
choke  must  have  to  keep  the  voltage  from  soar- 
ing when  the  external  load  is  removed. 

The  normal  current  rating  of  a power  supply 
usually  is  a round-number  value  chosen  on  the 
basis  of  the  transformers  and  chokes  on  hand 
or  available  from  the  catalog  of  a reliable  man- 
ufacturer. The  current  rating  of  a supply  to 
feed  a steady  load  such  as  a receiver,  a speech 
amplifier,  or  a continuously-operating  r-f  stage 
should  be  at  least  equal  to  the  steady  drain  of 
the  load.  However,  other  considerations  come 
into  play  in  choosing  the  current  rating  for  a 
keyed  amplifier,  an  amplifier  of  SSB  signals, 
or  a class  B modulator.  In  the  case  of  a sup- 
ply which  will  feed  an  intermittent  load  such 
as  these,  the  current  ratings  of  the  transform- 
ers and  chokes  may  be  less  than  the  maximum 
current  which  will  be  taken;  but  the  current 
ratings  of  the  rectifier  tubes  to  be  used  should 
be  at  least  equal  to  the  maximum  current  which 
will  be  taken.  That  is  to  say  that  300-ma. 
transformers  and  chokes  may  be  used  in  the 
supply  for  a modulator  whose  resting  current 
is  100  ma.  but  whose  maximum  current  at 
peak  signal  will  rise  to  500  ma.  However,  the 
rectifier  tubes  should  be  capable  of  handling 
the  full  500  ma. 


686  Power  Supplies 


THE  RADIO 


The  iron-core  components  of  a power  supply 
which  feeds  an-  intermittent  load  may  be  cho- 
sen on  the  basis  of  the  current  as  averaged 
over  a period  of  several  minutes,  since  it  is 
heating  effect  of  the  current  which  is  of  great- 
est importance  in  establishing  the  ratings  of 
such  components.  Since  iron-core  components 
have  a relatively  large  amount  of  thermal  in- 
ertia, the  effect  of  an  intermittent  heavy  cur- 
rent is  offset  to  an  extent  by  a key-up  period 
or  a period  of  low  modulation  in  the  case  of  a 
modulator.  However,  the  current  rating  of  a 
rectifier  tube  is  established  by  the  magnitude 
of  the  emission  available  from  the  filament  of 
the  tube;  the  maximum  emission  must  not  be 
exceeded  even  for  a short  period  or  the  recti- 
fier tube  will  be  damaged.  The  above  consider- 
ations are  predicated,  however,  on  the  assump- 
tion that  none  of  the  iron-core  components  will 
become  saturated  due  to  the  high  intermittent 
current  drain.  If  good-quality  components  of 
generous  weight  are  chosen,  saturation  will 
not  be  encountered. 

Voltage  The  general  subject  of  voltage 
Regulation  regulation  can  really  be  divided 
into  two  sub-problems,  which  dif- 
fer greatly  in  degree.  The  first,  and  more 
common,  problem  is  the  case  of  the  normal 
power  supply  for  a transmitter  modulator, 
where  the  current  drain  from  the  supply  may 
vary  over  a ratio  of  four  or  five  to  one.  In 
this  case  we  desire  to  keep  the  voltage  change 
under  this  varying  load  to  a matter  of  10  or 
15  per  cent  of  the  operating  voltage  under  full 
load.  This  is  a quite  different  problem  from 
the  design  of  a power  supply  to  deliver  some 
voltage  in  the  vicinity  of  250  volts  to  an  os- 
cillator which  requires  two  or  three  milliam- 
peres  of  plate  current;  but  in  this  latter  case 
the  voltage  delivered  to  the  oscillator  must  be 
constant  within  a few  volts  with  small  varia- 
tions in  oscillator  current  and  with  large  varia- 
tions in  the  a-c  line  voltage  which  feeds  the 
oscillator  power  supply.  An  additional  voltage 
regulation  problem,  intermediate  in  degree  be- 
tween the  other  two,  is  the  case  where  a load 
must  be  fed  with  10  to  100  watts  of  power  at 
a voltage  below  500  volts,  and  still  the  voltage 
variation  with  changes  in  load  and  changes 
in  a-c  line  voltage  must  be  held  to  a few 
volts  at  the  output  terminals. 

These  three  problems  are  solved  in  the  nor- 
mal type  of  installation  in  quite  different  man- 
ners. The  high-power  case  where  output  volt- 
age must  be  held  to  within  10  to  15  per  cent  is 
normally  solved  by  using  the  proper  value  of 
inductance  for  the  input  choke  and  proper 


value  of  bleeder  at  the  output  of  the  power 
supply.  The  calculations  ate  simple;  the  in- 
ductance of  the  power-supply  input  choke  at 
minimum  current  drain  from  the  supply  should 
be  equal  in  henries  to  the  load  resistance  on 
the  supply  ( at  minimum  load  current ) divided 
by  1000.  This  value  of  inductance  is  called 
the  critical  inductance  and  it  is  the  minimum 
value  of  inductance  which  will  keep  the  out- 
put voltage  from  soaring  in  a choke-input 
power  supply  with  minimum  load  upon  the 
output.  The  minimum  load  current  may  be 
that  due  to  the  bleeder  resistor  alone,  or  it 
may  be  due  to  the  bleeder  plus  the  minimum 
drain  of  the  modulator  or  amplifier  to  which 
the  supply  is  connected. 

The  low-voltage  low-current  supply,  such  as 
would  be  used  for  a v.f.o.  or  the  high-frequen- 
cy oscillator  in  a receiver,  usually  is  regu- 
lated with  the  aid  of  glow-discharge  gaseous- 
regulator  tubes.  These  regulators  are  usually 
called  "VR  tubes.”  Their  use  in  various  types 
of  power  supplies  is  discussed  in  Section  32-7. 
The  electronically-regulated  power  supply, 
such  as  is  used  in  the  10  to  100  watts  power 
output  range,  also  is  discussed  later  on  in  this 
chapter  and  examples  are  given. 

Ripple  The  ripple-voltage  limitation 

Considerofions  imposed  upon  a power  supply 
is  determined  by  the  load 
which  will  be  fed  by  the  supply.  The  tolerable 
ripple  voltage  from  a supply  may  vary  from 
perhaps  5 per  cent  for  a class  B or  class  C 
amplifier  which  is  to  be  used  for  a c-w  stage 
or  amplifier  of  an  FM  signal  down  to  a few 
hundredths  of  one  per  cent  for  the  plate-volt- 
age supply  to  a low-level  voltage  amplifier  in 
a speech  amplifier.  The  usual  value  of  ripple 
voltage  which  may  be  tolerated  in  the  supply 
for  the  majority  of  stages  of  a phone  transmit- 
ter is  between  0.1  and  2.0  per  cent. 

In  general  it  may  be  stated  that,  with  60- 
cycle  line  voltage  and  a single-phase  rectifier 
circuit,  a power  supply  for  the  usual  stages  in 
the  amateur  transmitter  will  be  of  the  choke- 
input  type  with  a single  pi-section  filter  fol- 
lowing the  input  choke.  A c-w  amplifier  or 
other  stage  which  will  tolerate  up  to  5 pet  cent 
ripple  may  be  fed  from  a power  supply  whose 
filter  consists  merely  of  an  adequate-size  in- 
put choke  and  a single  filter  capacitor. 

A power  supply  with  input  choke  and  a 
single  capacitor  also  will  serve  in  most  cases 
to  feed  a class  B modulator,  provided  the  out- 
put capacitor  in  the  supply  is  sufficiently  large. 
The  output  capacitor  in  this  case  must  be 
capable  of  storing  enough  energy  to  supply  the 


HANDBOOK 


Requirements  687 


peak-current  requirement  of  the  class  B tubes 
on  modulation  peaks.  The  output  capacitor  for 
such  a supply  normally  should  be  between  4 
/u,fd.  and  20  /rfd. 

Capacitances  larger  than  20  ^ifd.  involve  a 
high  initial  charging  current  when  the  supply 
is  first  turned  on,  so  that  an  unusually  large 
input  choke  should  be  used  ahead  of  the  ca- 
pacitor to  limit  the  peak-current  surge  through 
the  rectifier  tubes.  A capacitance  of  less  than 
4 /ifd.  may  reduce  the  power  output  capability 
of  a class  B modulator  when  it  is  passing  the 
lower  audio  frequencies,  and  in  addition  may 
superimpose  a low-frequency  "growl”  on  the 
output  signal.  This  growl  will  be  apparent  only 
when  the  supply  is  delivering  a relatively  high 
power  output;  it  will  not  be  present  when  mod- 
ulation is  at  a low  level. 

When  a stage  such  as  a low-level  audio  am- 
plifier requires  an  extremely  low  value  of  rip- 
ple voltage,  but  when  regulation  is  not  of  im- 
portance to  the  operation  of  the  stage,  the  high 
degree  of  filtering  usually  is  obtained  through 
the  use  of  a resistance-capacitance  filter. 
This  filter  usually  is  employed  in  addition  to 
the  choke-capacitor  filter  in  the  power  supply 
for  the  higher-level  stages,  but  in  some  cases 
when  the  supply  is  to  be  used  only  to  feed 
low-current  stages  the  entire  filter  of  the  pow- 
er supply  will  be  of  the  resistance-capacitance 
type.  Design  data  for  resistance-capacitance 
filters  is  given  in  a following  paragraph. 

When  a low-current  stage  requites  very  low 
tipple  in  addition  to  excellent  voltage  regula- 
tion, the  power  supply  filter  often  will  end 
with  one  or  mote  gaseous-type  voltage-regula- 
tor tubes.  These  VR  tubes  give  a high  degree 
of  filtering  in  addition  to  their  voltage-regu- 
lating action,  as  is  obvious  from  the  fact  that 
the  tubes  tend  to  hold  the  voltage  drop  across 
their  elements  to  a very  constant  value  regard- 
less of  the  current  passing  through  the  tube. 
The  VR  tube  is  quite  satisfactory  for  improv- 
ing both  the  regulation  and  ripple  character- 
istics of  a supply  when  the  current  drain  will 
not  exceed  25  to  35  ma.  depending  upon  the 
type  of  VR  tube.  Some  types  are  rated  at  a 
maximum  current  drain  of  30  ma.  while  others 
are  capable  of  passing  up  to  40  ma.  without 
damage.  In  any  event  the  minimum  current 
through  the  VR  tube  will  occur  when  the  as- 
sociated circuit  is  taking  maximum  current. 
This  minimum  current  requirement  is  5 ma, 
for  all  types  of  gaseous-type  voltage-regulator 

tubes. 

Other  types  of  voltage-regulation  systems, 
in  addition  to  VR  tubes,  exhibit  the  added 


TO  FULL-WAVE  RIPPLE  IN  TERMS  OF  C AT  FULL  LOAD 

RECTIFIER 

„ CAPACITANCE,  C PERCENT  Rl PPLE 


FIGURE  2 


TO  FULL-WAVE 
RECTIFIER 


RIPPLE  IN  TERMS  OF  LOAD  RESISTANCE 
LOAD,  OHMS  PERCENT  RIPPLE 

25000  (BLEEDER  only)  0.02 
15000  0.04 

10000  0.06 

5000  O.I 

3000  0.17 

2000  0.25 


FIGURE  3 


RIPPLE  IN  TERMS  OF  Cl  AND  C2  AT  FULL  LOAD 
Cl  C2  PERCENT  RIPPLE 

2 2 1.2 


25000 


FIGURE  4 


0,7 

0,25 

0,06 


characteristic  of  offering  a low  value  of  rip- 
ple across  their  output  terminals.  The  elec- 
tronic-type of  voltage-regulated  power  supply 
is  capable  of  delivering  an  extremely  small 
value  of  ripple  across  its  output  terminals, 
even  though  the  rectifier-filter  system  ahead 
of  the  regulator  delivers  a relatively  high 
value  of  ripple,  such  as  in  the  vicinity  of  5 to 
10  per  cent.  In  fact,  it  is  more  or  less  self 
evident  that  the  better  the  regulation  of  such 
a supply,  the  better  will  be  its  ripple  charac- 
teristic. It  must  be  remembered  that  the  ripple 
output  of  a voltage-regulated  power  supply  of 
any  type  will  rise  rapidly  when  the  load  upon 
the  supply  is  so  high  that  the  regulator  begins 
to  lose  control.  This  will  occur  in  a supply  of 
the  electronic  type  when  the  voltage  ahead  of 
the  series  regulator  tube  falls  below  a value 
equal  to  the  sum  of  the  minimum  drop  across 
the  tube  at  that  value  of  current,  plus  the  out- 
put voltage.  In  the  case  of  a shunt  regulator 
of  the  VR-tube  type,  the  regulating  effect  will 
fail  when  the  current  through  the  VR  tube 
falls  below  the  usual  minimum  value  of  about 
5 ma. 

Calculation  Although  figures  2,  3 and  4 
of  Ripple  give  the  value  of  ripple  volt- 
age for  several  more  or  less 
standard  types  of  filter  systems,  it  is  often  of 
value  to  be  able  to  calculate  the  value  of  rip- 
ple voltage  to  be  expected  with  a particular 

set  of  filter  components.  Fortunately,  the  ap- 
proximate ripple  percentage  for  normal  values 


688  Power  Supplies 


THE  RADIO 


12  HY 


TO  FULL-WAVE 
RECTIFIER 


Figure  5 

SAMPLE  FILTER  FOR 
CALCULATION  OF  RIPPLE 


of  filter  components  may  be  calculated  with 
the  aid  of  rather  simple  formulas.  In  the  two 
formulas  to  follow  it  is  assumed  that  the  line 
frequency  is  60  cycles  and  that  a full  wave  or 
a full-wave  bridge  rectifier  is  being  used.  For 
the  case  of  a single-section  choke-input  filter 
as  illustrated  in  figure  2,  or  for  the  tipple  at 
the  output  of  the  first  section  of  a two-section 
choke  input  filter  the  equation  is  as  follows, 
118 

Per  cent  ripple  = 

LC-1 

where  LC  is  the  product  of  the  input  choke 
inductance  in  henrys  (at  the  operating  current 
to  be  used)  and  the  capacitance  which  follows 
this  choke  expressed  in  microfarads. 

In  the  case  of  a two-section  filter,  the  per 
cent  ripple  at  the  output  of  the  first  section  is 
determined  by  the  above  formula.  Then  this 
percentage  is  multiplied  by  the  filter  reduction 
factor  of  the  following  section  of  filter.  This 
reduction  factor  is  determined  through  the  use 
of  the  following  formula: 

LC-1 

Filter  reduction  factor  = 

1.76 

Where  LC  again  is  the  product  of  the  in- 
ductance and  capacitance  of  the  filter  section. 
The  reduction  factor  will  turn  out  to  be  a deci- 
mal value,  which  is  then  multiplied  by  the  per- 
centage ripple  obtained  from  the  use  of  the 
preceding  formula. 

As  an  example,  take  the  case  of  the  filter 
diagrammed  in  figure  5.  The  LC  product  of  the 
first  section  is  16.  So  the  ripple  to  be  expected 
at  the  output  of  the  first  section  will  be:  118/ 
(16-1)  or  118/15,  which  gives  7.87  per  cent. 
Then  the  second  section,  with  an  LC  product 
of  48,  will  give  a reduction  factor  of:  1.76/ 
(48-1)  or  1.76/47  or  0.037.  Then  the  ripple 
percentage  at  the  output  of  the  total  filter  will 
be:  7.87  times  0.037  or  slightly  greater  than 
0.29  per  cent  ripple. 

Resistance-  In  many  applications  where  the 
Capacitance  current  drain  is  relatively  small. 
Filters  so  that  the  voltage  drop  across 

the  series  resistors  would  not  be 


excessive,  a filter  system  made  up  of  resistors 
and  capacitors  only  may  be  used  to  advantage. 
In  the  normal  case,  where  the  reactance  of  the 
shunting  capacitor  is  very  much  smaller  than 
the  resistance  of  the  load  fed  by  the  filter  sys- 
tem, the  ripple  reduction  per  section  is  equal 
to  1/  (25rRC).  In  terms  of  the  120-cycle  ripple 
from  a full-wave  rectifier  the  ripple-reduction 
factor  becomes:  1.33/RC  where  R is  expressed 
in  thousands  of  ohms  and  C in  microfarads. 
For  60-cycle  ripple  the  expression  is:  2.66/RC 
with  R and  C in  the  same  quantities  as  above. 


Filter  System  Many  persons  have  noticed. 
Resonance  particularly  when  using  an  in- 
put choke  followed  by  a 2-g.fd. 
first  filter  capacitor,  that  at  some  value  of 
load  current  the  power  supply  will  begin  to 
hum  excessively  and  the  rectifier  tubes  will 
tend  to  flicker  or  one  tube  will  seem  to  take 
all  the  load  while  the  other  tube  dims  out.  If 
the  power  supply  is  shut  off  and  then  again 
started,  it  may  be  the  other  tube  which  takes 
the  load;  or  first  one  tube  and  then  the  other 
will  take  the  load  as  the  current  drain  is 
varied.  This  condition,  as  well  as  other  less 
obvious  phenomena  such  as  a tendency  for  the 
first  filter  capacitor  to  break  down  regardless 
of  its  voltage  rating  or  for  rectifier  tubes  to 
have  short  life,  results  from  resonance  in  the 
filter  system  following  the  high-voltage  rec- 
tifier. 

The  condition  of  resonance  is  seldom  en- 
countered in  low-voltage  power  supplies  since 
the  capacitors  used  are  usually  high  enough 
so  that  resonance  does  not  occur.  But  in  high- 
voltage  power  supplies,  where  both  choke  in- 
ductance and  filter  capacitance  are  more  ex- 
pensive, the  condition  of  resonance  happens 
frequently.  The  product  of  inductance  and  ca- 
pacitance which  resonates  at  120  cycles  is 
1.77.  Thus  a 1-gfd.  capacitor  and  a 1.77  henry 
choke  will  resonate  at  120  cycles.  In  almost 
any  normal  case  the  LC  product  of  any  section 
in  the  filter  system  will  be  somewhat  greater 
than  1.77,  so  that  resonance  at  120  cycles  will 
seldom  take  place.  But  the  LC  product  for 
resonance  at  60  cycles  is  about  7.1.  This  is  a 
value  frequently  encountered  in  the  input  sec- 
tion of  a high-voltage  power  supply.  It  occurs 
with  a 2-g.fd.  capacitor  and  a choke  which  has 
3.55  henrys  of  inductance  at  some  current 
value.  With  a 2-/ifd.  filter  capacitor  following 
this  choke,  resonance  will  occur  at  the  current 
value  which  causes  the  inductance  of  the  choke 
to  be  3.55  henrys.  When  this  resonance  does 
occur,  one  rectifier  tube  (assuming  mercury- 


HANDBOOK 


Rectification  Circuits  689 


vapor  types)  will  dim  and  the  other  will  be- 
come much  brighter. 

Thus  we  see  that  we  must  avoid  the  LC 
products  of  1.77  and  7.1.  With  a swinging-type 
input  choke,  whose  inductance  varies  over  a 
5-to-l  range,  we  see  that  it  is  possible  for 
resonance  to  occur  at  60  cycles  at  a low  value 
of  current  drain,  and  then  for  resonance  to  oc- 
cur at  120  cycles  at  approximately  full  load 
on  the  power  supply.  Since  the  LC  product 
must  certainly  be  greater  than  1.77  for  satis- 
factory filtering  along  with  peak-current  limi- 
tation on  the  rectifier  tubes,  we  see  that  with 
a swinging-type  input  choke  the  LC  product 
must  still  be  greater  than  7.1  at  maximum  cur- 
rent drain  from  the  power  supply.  To  allow  a 
reasonable  factor  of  safety,  it  will  be  well  to 
keep  the  LC  product  at  maximum  current  drain 
above  the  value  10. 

From  the  above  we  see  that  we  can  just  get 
by  with  a ‘2-/ifd.  first  capacitor  following  the 
input  choke  when  this  choke  is  of  the  more- 
or-less  standard  5-25  henry  swinging  variety. 
Some  of  the  less  expensive  types  of  chokes 
swing  from  about  3 to  12  henrys,  so  we  see 
that  the  first  capacitor  must  be  greater  than  2 
gfd.  to  avoid  resonance  with  these  chokes.  A 
3-/afd.  capacitor  may  be  used  if  available,  but 
a standard  4-gfd.  unit  will  give  a greater  safe- 
ty factor. 

32-2  Rectification 
Circuits 

There  are  a large  number  of  rectifier  circuits 
that  may  be  used  in  the  power  supplies  for  sta- 
tion equipment.  But  the  simpler  circuits  are 
more  satisfactory  for  the  power  levels  up  to 
the  maximum  permitted  the  radio  amateur.  Fig- 
ure 6 shows  the  three  most  common  circuits 
used  in  power  supplies  for  amateur  equipment. 

Half-Wave  A half-wave  rectifier,  as  shown  in 
Rectifiers  figure  6A,  passes  one  half  of  the 
wave  of  each  cycle  of  the  alter- 
nating current  and  blocks  the  other  half.  The 
output  current  is  of  a pulsating  nature,  which 
can  be  smoothed  into  pure,  direct  current  by 
means  of  filter  circuits.  Flalf-wave  rectifiers 
produce  a pulsating  current  which  has  xero 
output  during  one-half  of  each  a-c  cycle;  this 
makes  it  difficult  to  filter  the  output  properly 
into  d.c.  and  also  to  secure  good  voltage  regu- 
lation for  varying  loads. 

Full-Wave  A full-wave  rectifier  consists  of 

Rectifiers  a pair  of  half-wave  rectifiers 

working  on  opposite  halves  of  the 
cycle,  connected  in  such  a manner  that  each 


n 


n 


Figure  6 

MOST  COMMON  RECTIFIER  CIRCUITS 

(A)  shows  a half-wave  rectifier  circuit,  (B)  is 
the  standard  full-wave  rectifier  circuit  used 
with  a dual  rectifier  or  two  rectifier  tubes, 
and  (C)  is  the  bridge  rectifier  circuit. 


half  of  the  rectified  a-c  wave  is  combined  in 
the  output  as  shown  in  figure  7.  This  pulsat- 
ing unidirectional  current  can  be  filtered  to 
any  desired  degree,  depending  upon  the  par- 
ticular application  for  which  the  power  supply 
is  designed. 

A full-wave  rectifier  may  consist  of  two 
plates  and  a filament,  either  in  a single  glass 
or  metal  envelope  for  low-voltage  rectification 
or  in  the  form  of  two  separate  tubes,  each  hav- 
ing a single  plate  and  filament  for  high-voltage 
rectification.  The  plates  are  conneaed  across 
the  high-voltage  a-c  power  transformer  wind- 
ing, as  shown  in  figure  6B.  The  power  trans- 
former is  for  the  purpose  of  transforming  the 
110-volt  a-c  line  supply  to  the  desired  second- 


690  Power  Supplies 


THE  RADIO 


TRANSFORMER 

SECONDARY 

VOLTAOE 


n A A 

:aaaaaa’ 


COMBINED  RECTIFIED 
VOLTAGE 
PLATES  N*  1 1 2 


''SMOOTHED  VOLTAGE 
AFTER  FIRST  SECTION 
OF  FILTER 


D.C.  VOLTAGE 
AVAILABLE  FOR 
RADIO  USE 


Figure  7 

FULL-WAVE  RECTIFICATION 

Showing  transformer  secondary  voltage,  the 
rectifi^  output  of  each  tube,  the  combined 
output  of  the  rectifier,  the  srrtoothed  voltage 
after  one  section  of  filter,  and  the  substan- 
tially pure  dx.  output  of  the  rectifier-filter 
after  additional  sections  of  filter. 


ary  a-c  voltages  for  filament  and  plate  sup- 
plies. The  transformer  delivers  alternating  cur- 
rent to  the  two  plates  of  the  rectifier  tube;  one 
of  these  plates  is  positive  at  any  instant  dur- 
ing which  the  other  is  negative.  The  center 
point  of  the  high-voltage  transformer  winding 
is  usually  grounded  and  is,  therefore,  at  zero 
voltage,  thereby  constituting  the  negative  B 
connection. 

While  one  plate  of  the  rectifier  tube  is  con- 
ducting, the  other  is  inoperative,  and  vice  ver- 
sa. The  output  voltages  from  the  rectifier  tubes 
are  connected  together  through  the  common 
rectifier  filament  circuit.  Thus  the  plates  al- 
ternately supply  pulsating  current  to  the  out- 
put (load)  circuit.  The  rectifier  tube  filaments 
or  cathodes  are  always  positive  in  polarity 
with  respect  to  the  plate  transformer  in  this 
type  of  circuit. 

The  output  current  pulsates  120  times  per 
second  for  a full-wave  rectifier  connected  to 
a 60-cycle  a-c  line  supply;  hence  the  output  of 
the  rectifier  must  pass  through  a filter  to 
smooth  the  pulsations  into  direct  current.  Fil- 
ters are  designed  to  select  or  reject  alternating 
currents;  those  most  commonly  used  in  a-c 
power  supplies  are  of  the  low-pass  type.  This 
means  that  pulsating  currents  which  have  a 
frequency  below  the  cutoff  frequency  of  the 
filter  will  pass  through  the  filter  to  the  load. 


Direct  current  can  be  considered  as  alternat- 
ing current  of  zero  frequency;  this  passes 
through  the  low-pass  filter.  The  120-cycle  pul- 
sations are  similar  to  alternating  current  in 
characteristic,  so  that  the  filter  must  be  de- 
signed to  have  a cutoff  at  a frequency  lower 
than  120  cycles  (for  a 60  cycle  a-c  supply). 

Bridge  The  bridge  rectifier  (figure  6C) 

Rectification  is  a type  of  full-wave  circuit 
in  which  four  rectifier  elements 
or  tubes  are  operated  from  a single  high-volt- 
age winding  on  the  power  transformer. 

While  twice  as  much  output  voltage  can  be 
obtained  from  a bridge  rectifier  as  from  a cen- 
ter-tapped circuit,  the  permissible  output  cur- 
rent is  only  one-half  as  great  for  a given  power 
transformer.  In  the  bridge  circuit,  four  recti- 
fiers and  three  filament  heating  transformer 
windings  are  needed,  as  against  two  rectifiers 
and  one  filament  winding  in  the  center-tapped 
full-wave  circuit.  In  a bridge  rectifier  circuit, 
the  inverse  peak  voltage  impressed  on  any  one 
rectifier  tube  is  halved,  which  means  that  tubes 
of  lower  peak  inverse  voltage  rating  may  be 
used  for  a given  voltage  output. 

32-3  Standard  Power 
Supply  Circuits 

Choke  input  is  shown  for  all  three  of  the 
standard  circuits  of  figure  6,  since  choke  input 
gives  the  best  utilization  of  rectifier-tube  and 
power  transformer  capability,  and  in  addition 
gives  much  better  regulation.  Where  greater 
output  voltage  is  a requirement,  where  the  load 
is  relatively  constant  so  that  regulation  is  not 
of  great  significance,  and  where  the  rectifier 
tubes  will  be  operated  well  within  their  peak- 
current  ratings,  the  capacitor-input  type  of  fil- 
ter may  be  used. 

The  capacitor-input  filter  gives  a no-load 
output  voltage  equal  approximately  to  the  peak 
voltage  being  applied  to  the  rectifier  tubes. 
At  full-load,  the  d-c  output  voltage  is  usually 
slightly  above  one-half  the  secondary  a-c  volt- 
age of  the  transformer,  with  the  normal  values 
of  capacitance  at  the  input  to  the  filter.  With 
large  values  of  input  capacitance,  the  output 
voltage  will  run  somewhat  higher  than  the 
r-m-s  secondary  voltage  applied  to  the  tubes, 
but  the  peak  current  flowing  through  the  rec- 
tifier tubes  will  be  many  times  as  great  as  the 
d<  output  current  of  the  power  supply.  The 
half-wave  rectifier  of  figure  6A  is  commonly 
used  with  capacitor  input  and  resistance-capaci- 
tance filter  as  a high-voltage  supply  for  a 
cathode-ray  tube.  In  this  case  the  current  drain 


HANDBOOK 


Standard  Circuits  691 


@ HALF  AND  FULL  VOLTAGE  BRIDGE  POWER  SUPPLY 


@ CENTER  TAPPED  METHOD  FOR  UNTAPPED  TRANSFORMERS 


© SPECIAL  FILTER  CIRCUIT  FOR  BRIDGE  RECTIFIER 


Figure  8 

SPECIAL  SINGLE-PHASE  RECTIFICATION  CIRCUITS 

A description  of  the  application  and  operation  of  each  of  these  special  circuits  is  given  in  the 

accompanying  text. 


is  very  small  so  that  the  peak-current  rating 
of  the  rectifier  tube  seldom  will  be  exceeded. 

The  circuit  of  figure  6B  is  most  commonly 
used  in  medium-voltage  power  supplies  since 
this  circuit  is  the  most  economical  of  filament 
transformers,  rectifier  tubes  and  sockets,  and 
space.  But  the  circuit  of  figure  6C,  commonly 
called  the  bridge  rectifier,  gives  better  trans- 
former utilization  so  that  the  circuit  is  most 
commonly  used  in  higher  powered  supplies. 
The  circuit  has  the  advantage  that  the  entire 
secondary  of  the  transformer  is  in  use  at  all 
times,  instead  of  each  side  being  used  alter- 
nately as  in  the  case  of  the  full-wave  rectifier. 
As  a point  of  interest,  the  current  flow  through 
the  secondary  of  the  plate  transformer  is  a sub- 
stantially pure  a-c  wave  as  a result  of  better 
transformer  utilization,  instead  of  the  pulsat- 
ing d-c  wave  through  each  half  of  the  power 
transformer  secondary  in  the  case  of  the  full- 
wave  rectifier. 

The  circuit  of  figure  6C  will  give  the  great- 
est value  of  output  power  for  a given  trans- 
former weight  and  cost  in  a single-phase  power 
supply  as  illustrated.  But  in  attempting  to 
bridge-rectify  the  whole  secondary  of  a trans- 


former designed  for  a full-wave  rectifier,  in 
order  to  obtain  doubled  output  voltage,  make 
sure  that  the  insulation  rating  of  the  trans- 
former to  be  used  is  adequate.  In  the  bridge 
rectifier  circuit  the  center  of  the  high-voltage 
winding  is  at  a d-c  potential  of  one-half  the 
total  voltage  output  from  the  rectifier.  In  a 
normal  full-wave  rectifier  the  center  of  the 
high-voltage  winding  is  grounded.  So  in  the 
bridge  rectifier  the  entire  high-voltage  second- 
ary of  the  transformer  is  subjected  to  twice 
the  peak-voltage  stress  that  would  exist  if  the 
same  transformer  were  used  in  a full-wave  rec- 
tifier. High-quality  full-wave  transformers  will 
withstand  bridge  operation  quite  satisfactorily 
so  long  as  the  total  output  voltage  from  the 
supply  is  less  than  perhaps  4500  volts.  But 
inexpensive  transformers,  whose  insulation 
is  just  sufficient  for  full-wave  operation,  will 
break  down  when  bridge  reaification  of  the 
entire  secondary  is  attempted. 

Special  Single-  Figure  8 shows  six  cir- 
Phose  Rectification  cults  which  may  prove 
Circuits  valuable  when  it  is  de- 

sired to  obtain  more  than 


692  Power  Supplies 


THE  RADIO 


I D C.— ► 


-O+Eo 


Eo  = i.as  Es 
Is  - 0.406  to.c. 

RIPPLE  FREQUENCY  « 6f 
R I PPL  E PE  RCENT  - 4. 2 
PEAK  INVERSE 
TUBE  VOLTACE 


-V.  2.09  Eo 
2.6S  Es 


-O+Eo 


Eo  ' 2.34  £$ 

Is  s 0.616  l0.c. 

RIPPLE  FREQUENCY  »6F 
RIPPLE  PERCENT*  4.2 
PEAK  INVERSE  •v,  1.0S  EO 
TUBE  VOLTACE  ^ 2.44  Es 


© 6-PHASE  BRIDGE 


Figure  9 
COMMON 
POLYPHASE- 
RECTIFICATION 
CIRCUITS 

These  circuits  are  used 
when  poiyphase  power  is 
avaiiabte  for  the  plate 
supply  of  a high-power 
transmitter.  The  circuit 
at  (B)  is  also  called  a 
three-phase  full-wave 
rectification  system.  The 
circuits  are  described  in 
the  accompanying  text. 


one  output  voltage  from  one  plate  transformer 
or  where  some  special  combination  of  voltages 
is  required.  Figure  8A  shows  a more  or  less 
common  method  for  obtaining  full  voltage  and 
half  voltage  from  a bridge  rectification  circuit. 
With  this  type  of  circuit  separate  input  chokes 
and  filter  systems  are  used  on  both  output 
voltages.  If  a transformer  designed  for  use 
with  a full-wave  rectifier  is  used  in  this  cir- 
cuit, the  current  drain  from  the  full-voltage 
tap  is  doubled  and  added  to  the  drain  from  the 
half-voltage  tap  to  determine  whether  the  rat- 
ing of  the  transformer  is  being  exceeded.  Thus 
if  the  transformer  is  rated  at  1250  volts  at  500 
ma.  it  will  be  permissible  to  pull  250  ma.  at 
2500  volts  with  no  drain  from  the  1250-volt 
tap,  or  the  drain  from  the  1250-volt  tap  may 
be  200  ma.  if  the  drain  from  the  2500-volt 
tap  is  150  ma.,  and  so  forth. 

Figure  8B  shows  a system  which  may  be 
convenient  for  obtaining  two  voltages  which 
are  not  in  a ratio  of  2 to  1 from  a bridge-type 
rectifier;  a transformer  with  taps  along  the 


winding  is  required  for  the  circuit  however. 
With  the  circuit  arrangement  shown  the  volt- 
age from  the  tap  will  be  greater  than  one-half 
the  voltage  at  the  top.  If  the  circuit  is  changed 
so  that  the  plates  of  the  two  rectifier  tubes 
are  connected  to  the  outside  of  the  winding 
instead  of  to  the  taps,  and  the  cathodes  of  the 
other  pair  are  connected  to  the  taps  instead 
of  to  the  outside,  the  total  voltage  ouput  of 
the  rectifier  will  be  the  same,  but  the  voltage 
at  the  tap  position  will  be  less  than  half  the 
top  voltage. 

An  interesting  variable-voltage  circuit  is 
shown  in  figure  8C.  The  arrangement  may  be 
used  to  increase  or  decrease  the  output  volt- 
age of  a conventional  power  supply,  as  repre- 
sented by  transformer  Ti,  by  adding  another 
filament  transformer  to  isolate  the  filament 
circuits  of  the  two  reaifier  tubes  and  adding 
another  plate  transformer  between  the  filaments 
of  the  two  tubes.  The  voltage  contribution  of 
the  added  transformer  T2  may  be  subtracted 
from  or  added  to  the  voltage  produced  by  Ti 


HANDBOOK 


Standard  Circuits  693 


simply  by  reversing  the  double-pole  double- 
throw switch  S.  A serious  disadvantage  of  this 
circuit  is  the  fact  that  the  entire  secondary 
winding  of  transformer  T2  must  be  insulated 
for  the  total  output  voltage  of  the  power  sup- 
ply. 

An  arrangement  for  operating  a full-wave 
rectifier  from  a plate  transformer  not  equipped 
with  a center  tap  is  shown  in  figure  8D.  The 
two  chokes  Li  must  have  high  inductance  rat- 
ings at  the  operating  current  of  the  plate  sup- 
ply to  hold  down  the  a-c  current  load  on  the 
secondary  of  the  transformer  since  the  total 
peak  voltage  output  of  the  plate  transformer 
is  impressed  across  the  chokes  alternately. 
However,  the  chokes  need  only  have  half  the 
current  rating  of  the  filter  choke  L2  for  a cer- 
tain current  drain  from  the  power  supply  since 
only  half  the  current  passes  through  each 
choke.  Also,  the  two  chokes  Li  act  as  input 
chokes  so  that  an  additional  swinging  choke 
is  not  required  for  such  a power  supply. 

A conventional  two-voltage  power  supply 
with  grounded  transformer  center  tap  is  shown 
in  figure  8E.  The  output  voltages  from  this 
circuit  are  separate  and  not  additive  as  in 
the  circuit  of  figure  8B.  Figure  8F  is  of  ad- 
vantage when  it  is  desired  to  operate  Class  B 
modulators  from  the  half-voltage  output  of  a 
bridge  power  supply  and  the  final  amplifier 
from  the  full  voltage  output.  Both  Li  and  Lz 
should  be  swinging  chokes  but  the  total  drain 
from  the  power  supply  passes  through  Li  while 
only  the  drain  of  the  final  amplifier  passes 
through  L2.  Capacitors  Ci  and  C2  need  be  rated 
only  half  the  maximum  output  voltage  of  the 
power  supply,  plus  the  usual  safety  factor. 
This  arrangement  is  also  of  advantage  in  hold- 
ing down  the  "key-up”  voltage  of  a c-w  trans- 
mitter since  both  Li  and  L2  ate  in  series,  and 
their  inductances  are  additive,  insofar  as  the 
"critical  inductance”  of  a choke-input  filter 
is  concerned.  If  4 fild.  capacitors  are  used  at 
both  Cl  and  C2  adequate  filter  will  be  obtained 
on  both  plate  supplies  for  hum-free  radiophone 
operation. 

Polyphase  It  is  usual  practice  in  commer- 

Rectificotion  cial  equipment  installations 

Circuits  when  the  power  drain  from  a 

plate  supply  is  to  be  greater 
than  about  one  kilowatt  to  use  a polyphase  rec- 
tification system.  Such  power  supplies  offer 
better  transformer  utilization,  less  ripple  out- 
put and  better  power  factor  in  the  load  placed 
upon  the  a-c  line.  However,  such  systems  re- 
quire a source  of  three-phase  (or  two-phase 
with  Scott  connection)  energy.  Several  of  the 


more  common  polyphase  rectification  circuits 
with  their  significant  characteristics  are 
shown  in  figure  9.  The  increase  in  ripple  fre- 
quency and  decrease  in  percentage  of  ripple 
is  apparent  from  the  figures  given  in  figure  9. 
The  circuit  of  figure  9C  gives  the  best  trans- 
former utilization  as  does  the  bridge  circuit 
in  the  single-phase  connection.  The  circuit 
has  the  further  advantage  that  there  is  no  aver- 
age d-c  flow  in  the  transformer,  so  that  three 
single-phase  transformers  may  be  used.  A tap 
at  half-voltage  may  be  taken  at  the  junction 
of  the  star  transformers,  but  there  will  be  d-c 
flow  in  the  transformer  secondaries  with  the 
power  supply  center  tap  in  use.  The  circuit  of 
figure  9A  has  the  disadvantage  that  there  is  an 
average  d-c  flow  in  each  of  the  windings. 
Rectifiers  Rectifying  elements  in  high-volt- 
age plate  supplies  ate  almost  in- 
variably electron  tubes  of  either  the  high-vacu- 
um  or  mercury-vapor  type,  although  selenium 
or  silicon  rectifier  stacks  containing  a large 
number  of  elements  are  often  used.  Low-volt- 
age high-current  supplies  may  use  argon  gas 
rectifiers  (Tungar  tubes),  selenium  rectifiers, 
or  other  types  of  dry-disc  rectification  ele- 
ments. The  xenon  rectifier  tubes  offer  some 
advantage  over  mercury-vapor  rectifiers  for 
high-voltage  applications  where  extreme  tem- 
perature ranges  are  likely  to  be  encountered. 
However,  such  rectifiers  (3B25  for  example) 
are  considerably  more  expensive  than  their 
mercury-vapor  counterparts. 

Peok  Inverse  Plate  In  an  a-c  circuit,  the  maxi- 
Voltage  ond  Peak  mum  peak  voltage  or  cur- 
Plate  Current  rent  is  V 2 or  1.41  times 

that  indicated  by  the  a-c 
meters  in  the  circuit.  The  meters  read  the  root- 
mean-square  (r.m.s.)  values,  which  are  the 
peak  values  divided  by  1.4l  for  a sine  wave. 

If  a potential  of  1,000  r.m.s.  volts  is  ob- 
tained from  a high-voltage  secondary  winding 
of  a transformer,  there  will  be  1,410-volts  peak 
potential  from  the  rectifier  plate  to  ground.  In 
a single-phase  supply  the  rectifier  tube  has 
this  voltage  impressed  on  it,  either  positively 
when  the  current  flows  or  "inverse”  when  the 
current  is  blocked  on  the  other  half-cycle.  The 
inverse  peak  voltage  which  the  tube  will  stand 
safely  is  used  as  a rating  for  rectifier  tubes. 
At  higher  voltages  the  tube  is  liable  to  arc 
back,  thereby  destroying  or  damaging  it.  The 
relations  between  peak  inverse  voltage,  total 
transformer  voltage  and  filter  output  voltage 
depend  upon  the  characteristics  of  the  filter 
and  rectifier  circuits  (whether  full-  or  half- 
wave, bridge,  single-phase  or  polyphase,  etc.). 


THE  RADIO 


694  Power  Supplies 


=0^ 


R- 


LINE  VOLTS-HEATER  VOLTS 
HEATER  AMPERES 


-fTMV-o+  @ 

LINE  RECTIF 

— 1 


ER 


II 1 — o- 


© 


SELENIUM 

LINE-RECTIFIER 


— © 


VOLTAGE  DOUBLER 
FULL-WAVE 


C24:'*’ 


O-F  ® 

VOLTAGE  DOUBLER 
HALF-WAVE 


-o- 


© 

SELENIUM 

RECTIFIER 

VOLTAGE 

QUAORUPLER 


Figure  10 

TRANSFORMERLESS  POWER-SUPPLY 
CIRCUITS 

Circuits  such  as  shown  above  are  also  fre- 
quently called  line-rectifier  circuits.  Seleni- 
um rectifiers,  vacuum  diodes,  or  gas  diodes 
may  be  used  as  the  rectifying  elements  in 
these  circuits. 


Rectifier  tubes  are  also  rated  in  terms  of 
peak  plate  current.  The  actual  direct  load  cur- 
rent which  can  be  drawn  from  a given  rectifier 
tube  or  tubes  depends  upon  the  type  of  filter 
circuit.  A full-wave  rectifier  with  capacitor  in- 
put passes  a peak  current  several  times  the 
direct  load  current. 

In  a filter  with  choke  input,  the  peak  cur- 
rent is  not  much  greater  than  the  load  current 
if  the  inductance  of  the  choke  is  fairly  high 
(assuming  full-wave  rectification). 

A full-wave  rectifier  with  two  rectifier  ele- 
ments requires  a transformer  which  delivers 
twice  as  much  a-c  voltage  as  would  be  the 
case  with  a half-wave  rectifier  or  bridge  rec- 
tifier. 


Mercury-Vapor  The  inexpensive  mercur’j-va- 
Rectifier  Tubes  por  type  of  rectifier  tube  is 
almost  universally  used  in 
the  high-voltage  plate  supplies  of  amateur  and 
commercial  transmitters.  Most  amateurs  are 
quite  familiar  with  the  use  of  these  tubes  but 
it  should  be  pointed  out  that  when  new  or 
long-unused  mercury-vapor  tubes  are  first 
placed  in  service,  the  filaments  should  be  op- 
erated at  normal  temperamre  for  approxi- 
mately twenty  minutes  before  plate  voltage 
is  applied,  in  order  to  remove  all  traces  of 
mercury  from  the  cathode  and  to  clear  any 
mercury  deposits  from  the  top  of  the  envelope. 
After  this  preliminary  warm-up  with  a new 
tube,  plate  voltage  may  be  applied  within  20 
to  30  seconds  after  the  time  the  filaments 
are  turned  on,  each  time  the  power  supply 
is  used.  If  plate  voltage  should  be  applied 
before  the  filament  is  brought  to  full  tem- 
perature, active  material  may  be  knocked  from 
the  oxide-coated  filament  and  the  life  of  the 
tube  will  be  greatly  shortened. 

Small  r-f  chokes  must  sometimes  be  con- 
nected in  series  with  the  plate  leads  of  mer- 
cury-vapor rectifier  tubes  in  order  to  prevent 
the  generation  of  radio-frequency  hash.  These 
r-f  chokes  must  be  wound  with  sufficiently 
heavy  wire  to  carry  the  load  current  and  must 
have  enough  inductance  to  attenuate  the  r-f 
parasitic  noise  current  to  prevent  it  from  flow- 
ing in  the  filter  supply  leads  and  then  being 
radiated  into  nearby  receivers.  Manufactured 
mercury- vapor  rectifier  hash  chokes  are  avail- 
able in  various  current  ratings  from  the  James 
Millen  Company  in  Malden,  Mass.,  and  from 
the  /.  W.  Miller  Company  in  Los  Angeles. 

When  mercury-vapor  rectifier  tubes  are  op- 
erated in  parallel  in  a power  supply,  small 
resistors  or  small  iron-core  choke  coils  should 
be  connected  in  series  with  the  plate  lead  of 
each  tube.  These  resistors  or  inductors  tend 
to  create  an  equal  division  of  plate  current  be- 
tween parallel  tubes  and  prevent  one  tube  from 
carrying  the  major  portion  of  the  current. 
When  high  vacuum  rectifiers  are  operated  in 
parallel,  these  chokes  or  resistors  are  not  re- 
quired. 

Transformerless  Figure  10  shows  a group  of 
Power  Supplies  five  different  types  of  trans- 
formerless power  supplies 
which  are  operated  directly  from  the  a-c  line. 
Circuits  of  the  general  type  are  normally  found 
in  a.c.-d.c.  receivers  but  may  be  used  in  low- 
powered  exciters  and  in  test  instruments.  When 
circuits  such  as  shown  in  (A)  and  (B)  are 
operated  directly  from  the  a-c  line,  the  rec- 


HANDBOOK 


Standard  Circuits  695 


tifier  element  simply  rectifies  the  line  voltage 
and  delivers  the  alternate  half  cycles  of  energy 
to  the  filter  network.  With  the  normal  type 
of  rectifier  tube,  load  currents  up  to  approxi- 
mately 7 5 ma.  may  be  employed.  The  d-c 
voltage  output  of  the  filter  will  be  slightly 
less  than  the  r-m-s  line  voltage,  depending 
upon  the  particular  type  of  rectifier  tube  em- 
ployed. With  the  introduction  of  the  miniature 
selenium  rectifier,  the  transformerless  power 
supply  has  become  a very  convenient  source 
of  moderate  voltage  at  currents  up  to  perhaps 
500  ma.  A number  of  advantages  are  offered 
by  the  selenium  rectifier  as  compared  to  the 
vacuum  tube  rectifier.  Outstanding  among 
these  are  the  factors  rhat  the  selenium  recti- 
fier operates  instantly,  and  that  it  requites  no 
heater  power  in  order  to  obtain  emission.  The 
amount  of  heat  developed  by  the  selenium  rec- 
tifier is  very  much  less  than  that  produced  by 
an  equivalent  vacuum-tube  type  of  rectifier. 

In  the  circuits  of  figure  10  (A),  (B)  and 
(C),  capacitors  G and  G should  be  rated 
at  approximately  150  volts  and  for  a normal 
degree  of  filtering  and  capacitance,  should  be 
between  15  to  60  /xfd.  In  the  circuit  of  figure 
lOD,  capacitor  Ci  should  be  rated  at  150 
volts  and  capacitor  Cz  should  be  rated  at  300 
volts.  In  the  circuit  of  figure  lOE,  capacitors 
Cl  and  Cl  should  be  rated  at  150  volts  and 
Cs  and  Ci  should  be  rated  at  300  volts. 

The  d-c  output  voltage  of  the  line  reaifier 
may  be  stabilized  by  means  of  a VR  tube. 
However,  due  to  the  unusually  low  internal 
resistance  of  the  selenium  rectifier,  transform- 
erless power  supplies  using  this  type  of  rec- 
tifying element  can  normally  be  expected  to 
give  very  good  regulation. 

Voltage-Doubler  Figures  IOC  and  lOD  illus- 
Circuits  trate  two  simple  voltage- 

doubler  circuits  which  will 
deliver  a d-c  output  voltage  equal  approximate- 
ly to  twice  the  r-m-s  value  of  the  power  line 
voltage.  The  no-load  d-c  output  voltage  is 
equal  to  2.82  times  the  r-m-s  line  voltage 
value.  At  high  current  levels,  the  output  volt- 
age will  be  slightly  under  twice  the  line  volt- 
age. The  circuit  of  figure  IOC  is  of  advantage 
when  the  lowest  level  of  ripple  is  required 
from  the  power  supply,  since  its  ripple  fre- 
quency is  equal  to  twice  the  line  frequency. 
The  circuit  of  figure  lOD  is  of  advantage  when 
it  is  desired  to  use  the  grounded  side  of  the 
a-c  line  in  a permanent  installation  as  the  re- 
turn circuit  for  the  power  supply.  However, 
with  the  circuit  of  figure  lOD  the  ripple  fre- 
quency is  the  same  as  the  a-c  line  frequency. 


©RELATIVE  LOAD  CURRENT, 

PERCENT  OF  FULL  LOAD. 

Figure  1 1 

THE  SELENIUM  RECTIFIER 
A — The  selenium  rectifier  Is  a semi-conductor 
stack  built  up  of  nickel  plated  aluminum 
discs  coated  on  one  side  with  selenium 
alloy. 

B — Rectifier  efficiency  is  high,  reaching  70% 
for  single  phase  service,  dropping  slightly 
at  high  current  densities. 


Voltage  The  circuit  of  figure  lOE  illus- 
Quodrupler  trates  a voltage  quadruplet  cir- 
cuit for  miniature  selenium  rec- 
tifiers. In  effect  this  circuit  is  equivalent  to 
two  voltage  doublers  of  the  type  shown  in  fig- 
ure lOD  with  their  outputs  connected  in  series. 
The  circuit  delivers  a d-c  output  voltage  under 
light  load  approximately  equal  to  four  times 
the  r-m-s  value  of  the  line  voltage.  The  no- 
load  d-c  output  voltage  delivered  by  the  quad- 
ruplet is  equal  to  5.66  times  the  r-m-s  line 
voltage  value  and  the  output  voltage  decreases 
rather  rapidly  as  the  load  current  is  increased. 

In  each  of  the  circuits  in  figure  10  where 
selenium  rectifiers  have  been  shown,  conven- 
tional high-vacuum  rectifiers  may  be  substi- 
tuted with  their  filaments  connected  in  series 
and  an  appropriate  value  of  the  line  resistor 
added  in  series  with  the  filament  string. 

32-4  Selenium  and 
Silicon  Rectifiers 

Selenium  rectifiers  are  characterized  by  long 

life,  dependability,  and  maintenance-free  op- 
eration under  severe  operating  conditions.  The 


696  Power  Supplies 


THE  RADIO 


Figure  12 

VOLTAGE  REGULATION  OF 
SELENIUM  CELL 

This  graph  applies  to  single  phase  tall  wave 
bridge,  and  center-tap  circuits  which  utilize 
both  halves  of  the  input  wave.  In  single  phase 
half  wave  circuits  the  regulation  will  be  poorer. 


selenium  rectifier  consists  of  a nickel-plated 
aluminum  base  plate  coated  with  selenium 
over  which  a low  temperature  alloy  is  sprayed. 
The  base  plate  serves  as  the  negative  electrode 
and  the  alloy  as  the  positive,  with  current 
flowing  readily  from  the  base  plate  to  the 
alloy  but  encountering  high  resistance  in  the 
opposite  direction  (figure  llA).  This  action 
results  in  effective  rectification  of  an  alternat- 
ing input  voltage  and  current  with  the  efficien- 
cy of  conversion  dependent  to  some  extent 
upon  the  ratio  of  the  resistance  in  the  con- 
ducting direction  to  that  of  the  blocking  di- 
rection. In  normal  power  applications  a ratio 
of  100  to  1 is  satisfactory;  however,  special 
applications  such  as  magnetic  amplifiers  often 
require  ratios  in  the  order  of  1000  to  1. 

The  basic  selenium  rectifier  cell  is  actually 
a diode  capable  of  half  wave  rectification. 
Since  many  applications  require  full  wave  rec- 
tification for  maximum  efficiency  and  mini- 
mum ripple,  a plurality  of  cells  in  series,  paral- 
lel, or  series-parallel  combinations  ate  stacked 
in  an  assembly. 

Selenium  rectifiers  are  operated  over  a wide 
range  of  voltages  and  currents.  Typical  appli- 
cations range  from  a few  volts  at  milliamperes 
of  current  to  thousands  of  amperes  at  rela- 
tively high  voltages. 

The  efficiency  of  high  quality  selenium  rec- 
tifiers is  high,  usually  in  the  order  of  90% 
in  three  phase  bridge  circuits  and  70%  in 
single  phase  bridge  circuits.  Of  particular  in- 
terest is  the  very  slight  decrease  in  efficiency 
even  at  high  current  overloads  (figure  IIB). 

Threshold  Voltage  A minimum  voltage  is  re- 
and  Aging  quired  to  permit  a selen- 

ium rectifier  to  conduct 
in  the  forward  direction.  This  voltage,  com- 
monly known  as  the  threshold  voltage,  pre- 
cludes the  use  of  selenium  rectifiers  at  ex- 


POSmvE  TERMINAL  CONTACT  NEGATIVE  TERMINAL 


I 

SILICON  CELL  LSPRING 

Figure  13 

THE  SILICON  CELL 

The  common  silicon  rectifier  is  a pressure 
contact  device  capable  of  operation  in  am- 
bient temperatures  as  high  as  ISO^C.  Heavy 
end  ferrules  that  fit  standard  fuse  clips  are 
large  enough  to  provide  ''heat  sink"  action. 
The  positive  ferrule  is  grooved  fo  provide 
polarity  identification  and  prevent  incorrect 
mounting. 


tremely  low  (less  than  one  volt)  applications. 
The  threshold  voltage  will  vary  with  tem- 
perature and  will  increase  with  a decrease  in 
temperature. 

Under  operating  conditions,  and  to  a lesser 
extent  when  idle,  the  selenium  rectifier  will 
age.  During  the  aging  period  the  forward 
resistance  will  gradually  increase,  stabilizing 
at  a new,  higher  value  after  about  one  year. 
This  aging  will  result  in  approximately  a 7% 
decrease  in  output  voltage. 

Voltage  The  selenium  rectifier  has  ex- 
Regulotion  tremely  low  internal  impedance 
which  exhibits  non-linear  charac- 
teristics with  respect  to  applied  voltage.  This 
results  in  good  voltage  regulation  even  at  large 
overload  currents.  Figure  12  shows  that  as  the 
load  is  varied  from  zero  to  300%  of  normal, 
the  output  voltage  will  change  about  10%. 
It  should  be  noted  that  because  of  non-linear 
characteristics,  the  voltage  drop  increases  rapid- 
ly below  50%  of  normal  load. 

Silicon  Of  all  recent  developments  in  the 
Rectifiers  field  of  semi-conductors,  silicon 
rectifiers  offer  the  most  promising 
range  of  applications;  from  extreme  cold  to 
high  temperature,  and  from  a few  watts  of 
output  power  to  very  high  voltage  and  cur- 
rents. Inherent  charaaeristics  of  silicon  allow 
junction  temperatures  in  the  order  of  200  °C 
before  the  material  exhibits  intrinsic  proper- 
ties. This  extends  the  operating  range  of  sili- 
con devices  beyond  that  of  any  other  efficient 
semi-conductor  and  the  excellent  thermal  range 
coupled  with  very  small  size  per  watt  of  out- 
put power  make  silicon  rectifiers  applicable 
where  other  rectifiers  were  previously  con- 
sidered impractical. 

Silicon  The  current  density  of  a sili- 

Current  Density  con  rectifier  is  very  high,  and 
on  present  designs  ranges 


HANDBOOK 


Mobile  Power  Supply  697 


from  600  to  900  amperes  per  square  inch  of 
effective  barrier  layer.  The  usable  current  den- 
sity depends  upon  the  general  construction  of 
the  unit  and  the  ability  of  the  heat  sink  to 
conduct  heat  from  the  crystal.  The  small  size 
of  the  crystal  is  illustrated  by  the  fact  that  a 
rectifier  rated  at  15  amperes  d.c.,  and  150 
amperes  peak  surge  current  has  a total  cell 
volume  of  only  .00023  inches.  Peak  currents 
are  extremely  critical  because  the  small  mass 
of  the  cell  will  heat  instantaneously  and  could 
teach  failure  temperatures  within  a time  lapse 
of  microseconds.  The  assembly  of  a typical 
silicon  cell  is  shown  in  figure  13. 

Operating  The  reverse  direction  of  a sili- 

Characteristics  con  rectifier  is  characterized 
by  extremely  high  resistance, 
up  to  10"  ohms  below  a critical  voltage  point. 
This  point  of  avalanche  voltage  is  the  region 
of  a sharp  break  in  the  resistance  curve,  fol- 
lowed by  rapidly  decreasing  resistance  (figure 
15A).  In  practice,  the  peak  inverse  working 
voltage  is  usually  set  at  least  20%  below  the 
avalanche  point  to  provide  a safety  factor. 

The  forward  direction,  or  direction  of  low 
resistance  determines  the  majority  of  power 
loss  within  the  semi-conductor  device.  Figure 
15B  shows  the  static  forward  current  charac- 
teristics versus  applied  voltage.  The  threshold 
voltage  is  about  0.6  volts. 

Since  the  forward  resistance  of  a semi-con- 
ductor is  very  low,  any  unbalance  between 
threshold  voltages  or  internal  voltage  drop 
would  cause  serious  unbalance  of  load  distri- 
bution and  ultimate  failure  of  the  overloaded 
section.  A small  resistance  should  therefore 
be  placed  in  series  with  each  half  wave  section 


Figure  14 

MINIATURE  SEMI-CONDUCTOR 
TYPE  RECTIFIER 

Raytheon  CK-777  power  rectifier  bolts  to 
chassis  to  gain  large  "heat  sink"  area.  Low 
internal  voltage  drop  and  high  efficiency 
permit  small  size  of  unit. 


operating  in  parallel  to  balance  the  load  cur- 
rents. 

Some  interesting  and  practical  semi-conduc- 
tor power  supplies  are  shown  in  figure  16. 
Remember  that  the  circuits  of  figure  16A  and 
B,  and  those  of  figure  10  are  "hot”  with  re- 
spect to  one  side  of  the  power  line. 

32-5  100  Watt  Mobile 

Power  Supply 

High  efficiency  and  compact  size  are  the 
most  important  factors  in  the  design  of  mobile 
power  supplies.  The  power  package  described 
in  this  section  meets  these  stringent  require- 


FORWARD  VOLTAGE  DROP 

VOLTS,  D.C. 

® 


Figure  15 

SILICON  RECTIFIER  CHARACTERISTICS 

A — Reverse  direction  of  silicon  rectifier  is  characterized  by  extremely  high  resistonce  up  to  point  of 
avalanche  voltage. 

B — Threshold  voltage  of  silicon  cell  is  about  0.6  volt.  Once  device  starts  conducting  the  current  in- 
creases exponentially  with  small  increments  of  voltage,  then  nearly  linearly  on  a very  steep  slope. 


698  Power  Supplies 


@ HIGH  CURRENT  SUPPLY  © 900  WATT  HIGH  VOLTAGE  SUPPLY 


Figure  16 

SEMI-CONDUCTOR  POWER  SUPPLIES 

A — Voltage  quadrupler  circuit.  If  point  *‘A“  is  taken  as  ground  instead  of  point  “B,"  supply  will  deliver 
530  volts  at  ISO  ma.  from  IIS  volt  a-c  line.  Supply  is  **hoi"  to  line, 

B — Voltage  tripler  delivering  32S  volts  at  4S0  ma.  Supply  is  **hot**  to  line. 

C — 900  watt  supply  for  sideband  service  may  be  made  from  two  voltage  quadruplets  working  in  series 
from  inexpensive  *'distribution^type"  transformer.  Sut^y  features  good  dynamic  voltage  regulation, 

PARTS  List: 

Dt^Sarkes  Tarzian  Model  150  selenium  cell  or  Model  M-SOO  silicon  cell, 

Df^Sarkes  Tarzian  Model  500  selenium  cell  or  Model  M-500  silicon  cell. 

Tr-^Power  distribution  transformer,  used  backwards.  230/460  primary,  1 15/230  secondary,  0,75  KVA. 
Chicago  PCB-24750. 


ments,  delivering  100  watts  at  various  voltages 
to  run  both  the  mobile  transmitter  and  re- 
ceiver. The  input  power  required  by  the  sup- 
ply is  almost  directly  proportional  to  the  out- 
put power  drain,  and  only  a small  amount  of 
power  is  wasted  to  actuate  the  vibrator  reeds. 

The  large  demand  for  efficient  and  power- 
ful mobile  radio  equipment  has  led  to  the 
development  of  new  heavy-duty,  vibrator-type 
power  supply  components;  these  ate  used  to 
advantage  in  this  unique  design. 

The  Split-Reed  The  new  split-reed  dual-inter- 
Vibrator  rupter  vibrator  overcomes  the 

power  capacity  limitations  of 
the  older  type  vibrators.  In  addition,  this  vi- 
brator permits  the  design  of  power  supplies 
requiring  no  component  changes  for  operation 
from  either  the  6-  or  12 -volt  d.c.  power  sys- 
tems with  which  most  automobiles  are 
equipped. 

Until  recently,  the  majority  of  vibrator  sup- 
plies have  been  designed  around  the  synchon- 
ous  type  vibrator.  A simplified  diagram  of 
such  a vibrator  is  shown  in  figure  18 A.  One 
set  of  vibrator  contacts  switches  the  battery 
current  alternately  through  two  opposed  pri- 
mary windings  of  a transformer,  thus  inducing 


a square-wave  a.c.  voltage  in  the  secondary. 
This  ax.  secondary  voltage  is  then  rectified  by  a 
second  set  of  contacts  on  the  vibrator  arma- 
ture. The  limitations  of  this  circuit  are  low 
power  handling  capacity  and  the  need  to  use 
a different  number  of  turns  on  the  transformer 
primary  when  the  battery  voltage  is  changed 
from  six  to  twelve  volts. 

In  the  split-reed  vibrator  ( figure  1 8B ) , two 
sets  of  double-throw  contacts  are  electrically 
isolated  from  each  other.  Each  set  has  much 
greater  current  carrying  capacity  than  the 
synchronous  vibrator.  However,  a power  trans- 
former having  two  center-tapped  primary 
windings  is  required.  One  set  of  contacts 
switches  the  d.c.  power  alternately  between 
halves  of  one  primary  winding,  and  the  other 
set  simultaneously  switches  the  other  winding. 
Therefore,  the  primaries  can  be  connected  in 
parallel  for  6-volt  operation,  or  in  series  for 
12-volt  operation.  No  wiring  changes  are 
needed  in  the  supply  if  the  battery  is  properly 
connected  to  the  power  input  terminals. 

The  Selenium  A selenium  rectifier  system 
Rectifier  System  is  employed  in  this  supply. 

Since  the  vibrator  contacts 
open  and  close  abruptly,  the  periodically  in- 


Figure  17 

100  WATT  MOBILE 
POWER  SUPPLY 

This  high  efficiency  mo- 
bile supply  features  6 or 
72  volt  input,  and  will 
completely  power  both 
the  mobile  transmitter 
and  receiver.  450,  300, 
and  240  volts  are  avail- 
able. At  right  is  control 
box  for  complete  mobile 
system.  Overall  size  of 
supply  is  only  6"x6"x6'’. 


terrupted  voltage  impressed  on  the  transform- 
er primary  has  practically  a square  waveform. 
The  secondary  voltage  also  has  nearly  a square 
waveform  as  a result  and  thus  the  peak  volt- 
age on  the  reaifiets  is  only  slightly  higher 
than  the  average  voltage  of  the  waveform. 
This  means  that  the  reaifiers  in  a vibrator- 
type  power  supply  can  be  operated  on  a square 
wave  voltage  close  to  their  maximum  peak 
inverse  voltage  rating,  instead  of  considerably 
below  this  rating,  as  when  a sine  wave  a.c. 
voltage  is  applied  to  them. 

Power  Supply  This  power  supply  has  three 
Circuit  high  voltage  sections  as  shown 

in  figure  19.  The  250  volt 
receiver  supply  is  shown  in  figure  19A,  the 
300  volt  low  voltage  transmitter  supply  in 
figure  19B,  and  the  450  volt  high  voltage 
transmitter  supply  is  shown  in  figure  19C. 


Note  that  each  rectifier  has  two  sections.  The 
part  numbers  correspond  to  the  markings 
shown  in  the  complete  schematic  of  figure 
20.  In  the  250  volt  circuit,  the  full  trans- 
former secondary  voltage  is  applied  to  a full 
wave  rectifier  consisting  of  half  of  rectifiers 
SRi  and  SR2.  These  two  rectifiers  form  a com- 
mon portion  of  all  three  reaifier  circuits.  The 
junction  of  the  two  rectifiers  is  grounded  and 
the  positive  250  volt  output  is  taken  from 
the  center-tap  transformer  lead.  This  is  the 
opposite  of  the  usual  full  wave  rectifier  cir- 
cuit. 

The  300  volt  d.c.  output  is  obtained  from 
a bridge  rectifier  circuit  (figure  19B)  con- 
sisting of  one-half  of  SRi  and  SRz  in  the 
ground  legs,  plus  SRa  in  the  two  bridge  legs 
from  which  the  positive  voltage  is  obtained. 

Another  bridge  rectifier  circuit  is  employed 
for  the  450  volt  d.c.  output,  again  with  the 


SYNCHRONOUS  VIBRATOR 

VIBRATOR  TRANSFORMER 


@ 


SPLIT  REED  DUAL  PRIMARY 

VIBRATOR  VIBRATOR  TRANSFORMER 


Figure  1 8 

SUPPLY  FEATURES  THE  NEW  SPLIT-REED  VIBRATOR  CAPABLE  OF  HANDLING 
100  WATTS  OF  POWER  AT  EITHER  6 OR  12  VOLTS 


A — Simplified  diagram  of  typical  synchronous  vibrator  and  rectifier  circuit. 

B — Simplified  diagram  of  typical  split-reed  vibrator  and  6-72  volt  d-c.  changeover  system. 


700  Power  Supplies 


THE  RADIO 


Figure  19 

SIMPLIFIED  CIRCUITS  OF  RECTIFIER  SYSTEMS  USED  IN  HIGH  EFFICIENCY 

MOBILE  SUPPLY 

A — Full  wave,  250  volt  recilHer. 

B — 300  volt  bridge  circuit  tapped  across  portion  of  the  high  voltage  winding, 

C — 450  volt  bridge  rectifier  circuit. 


two  grounded  legs  of  the  circuit  passing 
through  SRi  and  SRj.  The  other  two  sections 
of  these  rectifiers  form  the  legs  of  the  bridge 
from  which  the  positive  d.c.  voltage  is  taken. 

Since  there  is  little  ripple  voltage  in  the 
d.c.  output  from  the  rectifiers,  a single  40 
^fd.  filter  capacitor  on  the  300  volt  section 
of  the  supply  is  adequate.  Two  100  fiM.  ca- 
pacitors are  placed  in  series  across  the  450 
volt  section  and  a capacitor  input  filtet  is 
employed  for  the  low  voltage  section. 

The  Control  A "push-to-talk”  power  switch- 
Circuit  ing  circuit  is  included  in  the 

power  supply  so  that  the  250 
volt  output  may  be  applied  to  a mobile  re- 
ceiver or  converter.  This  is  accomplished  with 
one  pair  of  contacts  on  a d.p.d.t.  relay,  RYi, 
which  also  turns  on  the  300  volt  output  to  a 
transmitter  exciter  when  its  relay  coil  is  en- 
ergized. A second  d.p.d.t.  relay,  RYj,  turns 
on  the  450  volt  output  to  a transmitter  final 
amplifier  and  modulator  when  its  coil  is  ener- 
gized. It  also  provides  a power-reducing  fea- 
ture when  the  coil  is  not  energized  by  apply- 
ing 300  volts  to  the  450  volt  output  terminal 
and  250  volts  to  the  300  volt  output  terminal. 

Since  all  reaifiers  have  high  voltage  on 
them  continuously  when  the  supply  is  operat- 
ing, the  idling  current  flow  through  the  un- 
used rectifiers  when  receiving  is  reduced  to 
almost  zero  by  disconnecting  them  from  the 
filter  capacitors  and  bleeder  resistors  by  the 
action  of  the  relays. 

A special  20  volt  secondary  winding  on  the 
transformer  provides  transmitter  negative  bias 
through  rectifier  SR4  and  a 50  /rfd.  filter  ca- 
pacitor. 

Circuit  Details — In  this  power  supply. 

Low  Voltage  Section  changing  from  6-  to  12- 
volt  operation  is  accom- 
plished with  a 12-pin  Cinch- Jones  type  300 
power  plug  and  socket,  Pi  and  Ji,  respectively. 


A separate  low  voltage  input  cable  is  used 
for  each  voltage  as  shown  in  figure  20.  Leads 
from  Pi  to  the  main  power  relay  RYa  should 
be  made  as  short  as  possible  to  reduce  the 
voltage  drop.  This  is  particularly  important 
with  a six  volt  power  source,  where  a cable 
resistance  of  only  0.04  ohm  will  cause  a one 
volt  primary  drop  when  the  power  supply 
is  operating  at  full  load. 

The  six  volt  input  plug  conneas  the  alter- 
nate pairs  of  vibrator  contacts  and  the  halves 
of  the  transformer  primary  windings  in  paral- 
lel. These  units  are  connected  in  series  for 
twelve  volt  operation  as  shown  in  figure  18B. 
Six  volts  for  the  relay  coils  is  supplied  from 
the  power  cable  through  pin  6 on  Ji.  With  the 
twelve  volt  plug,  a voltage  dropping  resistor, 
Ri,  is  connected  in  series  with  the  relay  coils. 
If  the  power  supply  is  to  be  operated  exclu- 
sively from  the  higher  primary  voltage,  re- 
lays having  twelve  volt  coils  should  be  used, 
thus  eliminating  Ri. 

There  usually  is  sparking  at  the  vibrator 
contaas  so  various  capacitors  and  resistors 
are  incorporated  in  the  primary  circuit  to  sup- 
press any  radio  noise  generated  by  vibrator 
aaion.  These  components  are  placed  close 
to  the  pins  of  the  vibrator  socket. 

Component  The  heart  of  this  power  supply 
Ports  is  the  vibrator  transformer  de- 

signed specially  for  two-way  mo- 
bile radio  equipment.  It  is  readily  available 
from  many  of  the  3000  General  Electric  Co. 
mobile  radio  service  stations,  or  it  may  be 
ordered  from  the  address  given  in  the  parts 
list  of  figure  20. 

Power  Supply  All  components  except  the 
Construction  power  transformer  are  enclosed 
inside  a 6”x6”x6'’  aluminum 
utility  box  for  maximum  protection  against 
dirt  and  dust.  A sub-chassis  made  from  1/16- 


HANDBOOK 


Mobile  Power  Supply  701 


El — Vibrator,  dual  interrupter,  split-reed;  6 rott  toils,  lISjc.p.s.  reed  frequency,  7 pin  base.  Mallory  type 
1701,  Oak  V.6853,  Radiart  5722,  or  G.E.  A-7141S84-P5. 

Fi — 30  ampere  cartridge  fuse. 

F2 — 15  ampere  cartridge  fuse. 

Ji — 12  pin  male  chassis  plug.  Cinch-Jones  P-312- AB 

Pi — 72  pin  female  cable  socket.  Cinch-Jones  S-312-CCT 

Li — 7.5  henry,  100  ma.  Stancor  C-1421 

bg — 7 fxh.,  1 amp.  Ohmite  2-50  r-f  choke 

RYi,  RYg — D.P.D.T.  relay,  6 volt  coil.  Advance  MG/2c/6VD,  Potter  & Brumfield  MR-1 1D-6V,  Guardian 
200-60  Coil  and  200-2  assembly,  or  Ohmite  D05X-158T,  or  equiv. 

SRi,  5Rg — Two  section  selenium  rectifier,  150  ma.,  380  volts  peak  inverse  per  section,  connected  for 
doubler  service.  G.E.  A-7144141-P2,  or  two  Federal  1005-A  rectifiers  in  series  for  each  leg  (8  in  all). 

SR3 — Two  section  selenium  rectifier,  150  ma.,  380  volts  peak  inverse  per  section,  connected  for  center-tap 
service.  G.E.  A-7144141-P1 , or  two  Federal  1005-A  rectifiers  with  red  terminals  connected  together. 

5Ri — 750  ma.,  64  volt  peak  inverse  selenium  rectifier.  G.  E.  A-7140806-P1  or  Federal  7075. 

Ti — Vibrator-type  power  transformer.  Dual  center-tapped  6 volt  primaries.  Secondaries:  420  volts  c.i. 
with  750  volt  taps,  300  ma.,  and  20  volt,  150  ma.  bias  voltage  winding.  G.E.  B-7486449-P1. 

TSi — 8 terminal  barrier  strip.  Cinch-Jones  8-141-Y. 

Note:  General  Electric  parts  may  be  ordered  from:  D.  S.  Clark,  G.E.  Co.,  Product  Service  Renewal  Parts 
Section  of  Communication  Products,  509  Kent  St.,  Utica,  N.Y.  Current  prices  plus  shipping  charges 
are:  Ti,  $14.70;  Ei,  $5.75;  SR„  SRg,  SRs,  $7,75  each;  SR^,  $1.40. 


inch  thick  aluminum  is  cut  as  shown  in  fig- 
ures 21  and  22.  A cut-out  in  one  cornet  is 
necessary  to  clear  the  lower  portion  of  the 
power  transformer.  All  chassis  drilling  and 
the  cutouts  for  the  power  transformer,  power 
input  plug,  and  output  terminal  strip  should 
be  completed  before  the  chassis  is  permanent- 


ly fastened  in  place  with  small  aluminum 
angle  brackets,  two  inches  from  the  bottom  of 
the  box. 

One  electrolytic  capacitor,  Ci,  is  mounted 
on  an  insulated  plate  furnished  with  the  ca- 
pacitor. When  twisting  the  locking  lugs,  make 
sure  that  they  do  not  come  near  the  metal 


Figure  21 
UNDER-CHASSIS 
VIEW  OF 
POWER  SUPPLY 

The  vibrator  socket  is 
hidden  by  the  resistors 
and  disc  ceramic  capaci- 
tors of  the  hash  filter. 
K-F  choke  Li  and  the 
25  iifd.  capacitors  are 
above  these  parts  on  the 
wall  of  the  case.  Resis- 
tor Ri  is  mounted  to  the 
wall,  with  SRif  SRtf  and 
SRi  in  line  below  it. 


chassis.  The  can  of  this  capacitor  is  more  than 
200  volts  "above  ground”  and  should  be  in- 
sulated with  a fibre  sleeve. 

The  large  seven  pin  socket  into  which  the 
vibrator  plugs  should  have  two  soldering  lugs 
placed  under  each  mounting  bolt  for  the  by- 
pass capacitors  which  are  later  wired  to  the 
socket  pins.  If  the  power  supply  is  to  be 
mounted  with  the  chassis  in  a vertical  plane, 
pins  1 and  4 of  the  vibrator  socket  should  be 
in  a vertical  plane.  The  smaller  components 
below  the  chassis  should  not  be  installed  until 
parts  above  the  chassis  have  been  assembled 
and  wired. 


The  transformer  primary  leads  are  wired 
to  the  input  plug  and  vibrator  socket  with 
shortest  possible  leads  before  other  parts  are 
wired.  Placement  of  the  under-chassis  parts 
may  be  seen  in  figure  21.  Layout  of  the  parts 
is  not  critical. 

Testing  the  After  the  wiring  has  been 
Supply  checked,  the  power  supply  should 
be  checked  at  half  input  voltage, 
but  with  full  voltage  applied  to  the  vibrator 
coil.  This  is  done  by  temporarily  removing 
the  jumper  between  pins  7 and  8 on  the 
twelve  volt  power  plug,  Pi.  Pin  8 is  then 


Figure  22 
TOP  VIEW  OF 
POWER  SUPPLY 
INTERIOR 

Relays  RYt  and  RYt  are 
at  center  above  choke  Li. 
Power  transformer  Ti  is 
mounted  on  end  of  cab- 
inet. Across  top  of 
chassis  (left  to  right)  are 
vibrator  and  filter  ca- 
pacitors 


Transistorized  Power  Supply  703 


FROM 

BATrERY 


3 WIRE  CABLE 
6 V.-  ite 
1 2 V.  - ♦ 1 2 


2 WIRE  CABLE 


POWER 

RELAY 

» 6 FOR  6 V. 
W10  FOR  12  V. 

P'l 

MOBILE 

AND 

FUSE 

SUPPLY 

TSi 

5 WIRE 
► HI-VOLTACe 
A6IAS  CABLE 


@ 


12  3 4 1 2 3 4 

JSTO  RECEIVER  J4  TO  TRANSMITTER 


ft  VOLT  POWER  CABLE 


12  VOLT  POWER  CABLE 


Figure  24 

CABLE  HARNESS,  MAIN  POWER 
RELAY,  RY3,  AND  REAR  OF 
POWER  CONTROL  BOX 

High  voltage  for  the  receiver,  plus  the  control 
switch  circuits  for  RYi  and  RYi,  are  brought 
into  the  control  box  from  the  power  supply 
through  J3. 

High  voltage  leads  for  the  transmitter  are 
run  directly  from  the  supply,  but  the  trans- 
mitter heater  power  and  ^'transmit-receive" 
control  circuit  is  run  to  the  transmitter 
through  control  box  circuit  Ji,  (figure  2ZA). 


any  hash  still  present  during  reception.  Every 
experienced  "mobileer"  will  agree  that  noise 
in  each  mobile  installation  usually  must  be 
eliminated  on  a "search  and  filter”  basis. 


rnPYi  TO  HEATERS, 
TOKT3  ft  VOLTS 

FROM  Pi 
(F/(:.20) 


COIL 
(F/G.  20) 


1 2 3 4 S 6 7 & 


Y Y 1 

M 

Y 

( Y Y 

r 

TO  R 
COI  L 
(F/G. 

Y3 

20) 

nz 

1 1 

1 

lEATEAS, 

tOLTS 

M Pi 
G.20) 

TO  F 
12  V 
FRO 
(A/( 

© 


Figure  23 

CONTROL  WIRING 

A — Block  diagram  of  suggested  power  and 
control  cables  and  switching  system  for 
mobile  power  supply. 

B — Schematic  diagram  of  suggested  control 
box  including  power  plugs  for  chonging 
heaters  for  6 or  12  vott  operation.  Si  is 
"transmit-receive  switch/^  Sr  is  ''high-iow" 
switch,  and  Sj  is  main  power  switch  for 
RYr. 

C — 6 and  12  volt  power  cables. 


jumped  to  pin  9-  The  cable  is  attached  to  Ji  on 
the  supply  and  to  a six  volt  power  source.  Ap- 
proximately half  the  rated  voltage  should  be 
measured  at  the  output  terminal  strip  if  the 
supply  has  been  properly  wired. 

Replace  the  original  connections  on  the 
power  plug  and  test  the  supply  with  full 
input  voltage.  A 2500  ohm,  100  watt  resistor, 
or  four  25  watt  115  volt  lamp  bulbs  in  series 
make  a good  load  resistance.  The  output  volt- 
ages should  measure  close  to  450,  300,  and 
240  volts  under  load.  Additional  .02  pfd.  by- 
pass capacitors  at  RYs,  the  output  terminal 
strip,  and  the  control  box  should  eliminate 


Installation  This  power  supply  may  be  op- 
in  the  Car  erated  in  conjunction  with  the 
suggested  configuration  shown  in 
figure  23A.  Note  that  a separate  cable  is 
recommended  for  the  heater  power  circuit  to 
reduce  vibrator  hash  pickup,  and  to  minimize 
heater  voltage  variations  when  the  supply  is 
switched  from  receive  to  transmit. 

Provision  has  been  made  for  changing  the 
mobile  receiver  and  transmitter  power  circuits 
for  either  six  or  twelve  volt  operation  as  shown 
in  the  schematic  of  the  control  box  in  figure 
23B,  An  eight  contact  plug  and  socket  auto- 
matically make  the  proper  connections  when 
the  six  or  twelve  volt  plug  is  attached. 

32-6  Transistorized 
Power  Supplies 

The  vibrator  type  of  mobile  supply  achieves 
an  overall  efficiency  in  the  neighborhood  of 
70%.  The  vibrator  may  be  thought  of  as  a 
mechanical  switch  reversing  the  polarity  of 
the  primary  source  at  a repetition  rate  of  120 
transfers  per  second.  The  switch  is  actuated  by 
a magnetic  coil  and  breaker  circuit  requiring 
appreciable  power  which  must  be  supplied  by 
the  primary  source. 

One  of  the  principal  applications  of  the 
transistor  is  in  switching  circuits.  The  tran- 
sistor may  be  switched  from  an  "off”  con- 
dirion  to  an  "on”  condition  with  but  the  ap- 


704  Power  Supplies 


THE  RADIO 


Figure  25 

TRANSISTORS  CAN  REPLACE  VIBRATOR  IN  MOBILE  POWER  SUPPLY  SYSTEM 

A — Typical  vibrator  circuit. 

B — Vibrator  cart  be  represented  by  two  single-pole  single  throw  switches,  or  transistors. 

C — Push-pull  square  wave  "oscillator''  is  driven  by  special  feedback  windings  on  power  transformer. 
D — Addition  of  bias  in  base-emitter  circuit  results  in  oscillator  capable  of  starting  under  full  load. 
a 


plication  of  a minute  exciting  signal.  W'hen 
the  transistor  is  nonconductive  it  may  be  con- 
sidered to  be  an  open  circuit.  When  it  is  in 
a conductive  state,  the  internal  resistance  is 
very  low.  Two  transistors  properly  connected, 
therefore,  can  replace  the  single  pole,  double 
throw  mechanical  switch  representing  the  vi- 
brator. The  transistor  switching  action  is  many 
times  faster  than  that  of  the  mechanical  vi- 
brator and  the  transistor  can  switch  an  ap- 
preciable amount  of  power.  Efficiencies  in  the 
neighborhood  of  95%  can  be  obtained  with  28 
volt  primary-type  transistor  power  supplies, 
permitting  great  savings  in  primary  power 
over  conventional  vibrators  and  dynamotors. 

Transistor  The  transistor  operation  resembles 

Operation  a magnetically  coupled  multi-vi- 

brator, or  an  audio  frequency 
push-pull  square  wave  oscillator  (figure  25C). 
A special  feedback  winding  on  the  power  trans- 


TIME  ► 

Figure  26 

BASE-COLLECTOR  WAVEFORM  OF 
SWITCHING  CIRCUIT, 

FOR  12  VOLT  CIRCUIT 

Square  waveshape  produces  almost  ideal 
switchirrg  action.  Small  "spike'*  on  leading 
edge  of  pulses  may  be  reduced  by  proper 
transformer  design. 


former  provides  180  degree  phase  shift  volt- 
age necessary  to  maintain  oscillation.  In  this 
application  the  transistors  are  operated  as  on- 
off  switches;  i.e.,  they  are  either  completing 
the  circuit  or  opening  it.  The  oscillator  output 
voltage  is  a square  wave  having  a frequency 
that  is  dependent  upon  the  driving  voltage, 
the  primary  inductance  of  the  power  trans- 
former, and  upon  the  peak  collector  current 
drawn  by  the  conducting  transistor.  Changes 
in  transformer  turns,  core  area,  core  material, 
and  feedback  turns  ratio  have  an  effect  on  the 
frequency  of  oscillation.  Frequencies  in  com- 
mon use  are  in  the  range  of  120  c.p.s.  to 
3,500  c.p.s. 

The  power  consumed  by  the  transistors  is 
relatively  independent  of  load.  Loading  the 
oscillator  causes  an  increase  in  input  current 
that  is  sufficient  to  supply  the  required  power 
to  the  load  and  the  additional  losses  in  the 
transformer  windings.  Thus,  the  overall  effi- 
ciency actually  increases  with  load  and  is  great- 
est at  the  heaviest  load  the  oscillator  will  sup- 
ply. A result  of  this  is  that  an  increase  in  load 
produces  very  little  extra  heating  of  the  tran- 
sistors. This  feature  means  that  it  is  impossi- 
ble to  burn  out  the  transistors  in  the  event  of 
a shorted  load  since  the  switching  action  mere- 
ly stops. 

Transisl-or  The  power  capability  of  the 
Power  Rating  transistor  is  limited  by  the 
amount  of  heat  created  by  the 
current  flow  through  the  internal  resistance 
of  the  transistor.  When  the  transistor  is  con- 
ducting the  internal  resistance  is  extremely 
low  and  little  heat  is  generated  by  current 
flow.  Conversely,  when  the  transistor  is  in  a 
cut-off  condition  the  internal  resistance  is  very 
high  and  the  current  flow  is  extremely  small. 
Thus,  in  both  the  "on”  and  "off”  conditions 


HANDBOOK 


Transistorized  Power  Supplies  705 


2N278 


2N278  OR  2N301A 


Ti — Chicago  DCT~2  transistor  transformer 
Ti — Chicago  DCT-1  transistor  transformer 
Di-Ds — Sarkes-Tarzian  M~500  silicon  rectifier  or  equivalent. 


the  transistor  dissipates  a minimum  of  power. 
The  important  portion  of  the  operating  cycle 
is  that  portion  when  the  actual  switching  from 
one  transistor  to  the  other  occurs,  as  this  is 
the  time  during  which  the  transistor  may  be 
passing  through  the  region  of  high  dissipation. 
The  greater  the  rate  of  switching,  in  general, 
the  faster  will  be  the  rise  time  of  the  square 
wave  (figure  26)  and  the  lower  will  be  the 
internal  losses  of  the  transistor.  The  average 
transistor  can  switch  about  eight  times  the 
power  rating  of  class  A operation  of  the 
unit.  Two  switching  transistors  having  5 watt 
class  A power  output  rating  can  therefore 
switch  80  watts  of  power  when  working  at 
optimum  switching  frequency. 

Self-Starting  The  transistor  supply  shown  in 
Oscillators  figure  25C  is  impractical  be- 
cause oscillations  will  not  start 
under  load.  Base  current  of  the  proper  po- 
larity has  to  be  momentarily  introduced  into 
the  base-emitter  circuit  before  oscillation  will 
start  and  sustain  itself.  The  addition  of  a bias 
resistor  (figure  25D)  to  the  circuit  results 


in  an  oscillator  that  is  capable  of  starting 
under  full  load.  Ri  is  usually  of  the  order  of 
10-50  ohms  while  R2  is  adjusted  so  that  ap- 
proximately 100  milliamperes  flows  through 
the  circuit. 

The  current  drawn  from  the  battery  by  this 
network  flows  through  R^  and  then  divides 
between  Ri  and  the  input  resistances  of  the 
two  transistors.  The  current  flowing  in  the 
emitter-base  circuit  depends  upon  the  value 
of  input  resistance.  The  induced  voltage  across 
the  feedback  winding  of  the  transformer  is  a 
square  wave  of  such  polarity  that  it  forward- 
biases  the  emitter-base  diode  of  the  transistor 
that  is  starting  to  conduct  collector  current, 
and  reverse-biases  the  other  transistor.  The 
forward-biased  transistor  will  have  a very  low 
input  impedance,  while  the  input  impedance 
of  the  reversed-biased  transistor  will  be  quite 
high.  Thus,  most  of  the  starting  current  drained 
from  the  primary  power  source  will  flow 
in  Ri  and  the  base-emitter  circuit  of  the  for- 
ward-biased transistor  and  very  little  in  the 
other  transistor.  It  can  be  seen  that  Ri  must  not 
be  too  low  in  comparison  to  the  input  re- 


706  Power  Supplies 


THE  RADIO 


Figure  28 

35  WATT  TRANSISTOR 
POWER  SUPPLY 

Two  2N301A  power  transistors  are  used  in 
this  midget  supply.  Transistors  are  mounted 
on  sand^  portion  of  chassis  deck  which  acts 
as  ''heat  sink."  See  text  for  details. 


sistance  of  the  conducting  transistor,  or  else 
it  will  shunt  too  much  current  from  the  tran- 
sistor. When  switching  takes  place,  the  trans- 
former polarities  reverse  and  the  additional 
current  now  flows  in  the  base-emitter  circuit 
of  the  other  transistor. 

The  Power  The  power  transformer  in  a 
Transformer  transistor-type  supply  is  designed 
to  reach  a state  of  maximum  flux 
density  (saturation)  at  the  point  of  maximum 
transistor  conductance.  When  this  state  is 
reached  the  flux  density  drops  to  zero  and 
reduces  the  feedback  voltage  developed  in  the 
base  winding  to  zero.  The  flux  then  reverses 


because  there  is  no  conducting  transistor  to 
sustain  the  magnetizing  current.  This  change 
of  flux  induces  a voltage  of  the  opposite  po- 
larity in  the  transformer.  This  voltage  turns 
the  first  transistor  off  and  holds  the  second 
transistor  on.  The  transistor  instantly  reaches 
a state  of  maximum  conduction,  producing  a 
state  of  saturation  in  the  transformer.  This 
action  repeats  itself  at  a very  fast  rate.  Switch- 
ing time  is  of  the  order  of  5 to  10  micro- 
seconds, and  saturation  time  is  perhaps  200 
to  2,000  microseconds.  The  collector  wave- 
form of  a typical  transistor  supply  is  shown 
in  figure  26.  The  rise  time  of  the  wave  is 
about  5 microseconds,  and  the  saturation  time 
is  500  microseconds.  The  small  "spike”  at 
the  leading  edge  of  the  pulse  has  an  ampli- 
tude of  about  IVl  volts  and  is  a product  of 
switching  transients  caused  by  the  primary 
leakage  reactance  of  the  transformer.  Proper 
transformer  design  can  reduce  this  "spike”  to 
a minimum  value.  An  excessively  large  "spike” 
can  puncture  the  transistor  junction  and  ruin 
the  unit. 

32-7  Two 

Transistorized 
Mobile  Supplies 

The  new  Chicago-Standard  Transformer  Co. 
series  of  power  transformers  designed  to  work 
in  transistor-type  power  supplies  permits  the 
amateur  and  experimenter  to  construct  ef- 
ficient mobile  power  supplies  at  a fraction 
of  their  former  price.  Described  in  this  section 
are  two  power  supplies  designed  around  these 
efficient  transformers.  The  smaller  supply  de- 
livers 35  watts  (275  volts  at  125  milliam- 
peres)  and  the  larger  supply  delivers  85  watts 
(500  volts  at  125  milliamperes  and  250  volts 


Figure  29 

SCHEMATIC,  TRANSISTOR 
POWER  SUPPLY  FOR 
12  VOLT  AUTOMOTIVE 
SYSTEM 

Ti — Transistor  power  transformer. 
12  volt  primary,  to  provide  275 
volts  at  725  mo.  Chicago  Stand- 
ard OCT- 7. 

D j-Di — Sarkes-T arzian  silicon  recti- 
fier, type  M-SOO,  or  equivalent. 


2N301A 


2.Z  K 
. 1W 


SOllF 
50  V. 


47 

TW  Bffr 
-n/W^ 


2N301A 


^1, 


£i- 


Di  Da 

-M  M 


0075 
1.6  KV. 


-H  H 

Ds  D4 


2QAJFI 

450 


ifT  25  Kf 
V-l  5W1 


275  V. 
AT125MA, 


6-  12  V.  BATTERY  +6 


HANDBOOK 


Components  707 


at  50  milliamperes,  simultaneously).  Both 
power  units  operate  from  a 12  volt  primary 
source. 

The  35  Watt  The  35  watt  power  unit  uses 
Supply  two  inexpensive  RCA  2N301A 

P-N-P-type  power  transistors 
for  the  switching  elements  and  four  silicon 
diodes  for  the  high  voltage  rectifiers.  The 
complete  schematic  is  shown  in  figure  29. 
Because  of  the  relatively  high  switching  fre- 
quency only  a single  20  gfd.  filter  capacitor 
is  required  to  provide  pure  d.c. 

Regulation  of  the  supply  is  remarkably  good. 
No  load  voltage  is  310  volts,  dropping  to 
275  volts  at  maximum  current  drain  of  125 
milliamperes. 

The  complete  power  package  is  built  upon 
an  aluminum  chassis-box  measuring  5V4”x 
3”x2”.  Paint  is  removed  from  the  center  por- 
tion of  the  box  to  form  a simple  heat  sink 
for  the  transistors.  The  box  therefore  conducts 
heat  away  from  the  collector  elements  of  the 
transistors.  The  collector  of  the  transistor  is 
the  metal  case  terminal  and  in  this  circuit  is 
returned  to  the  negative  terminal  of  the  pri- 
mary supply.  If  the  negative  of  the  automobile 
battery  is  grounded  to  the  frame  of  the  car 
the  case  of  the  transistor  may  be  directly 
grounded  to  the  unpainted  area  of  the  chassis. 
If  the  positive  terminal  of  the  car  battery  is 
grounded  it  is  necessary  to  electrically  insulate 
the  transistor  from  the  aluminum  chassis,  yet 
at  the  same  time  permit  a low  thermal  barrier 
to  exist  between  the  transistor  case  and  the 
power  supply  chassis.  A simple  method  of 
accomplishing  this  is  to  insert  a thin  mica 
sheet  between  the  transistor  and  the  chassis. 
"Two  Mil”  •(0.002”)  mica  washers  for  tran- 
sistors are  available  at  many  large  radio  sup- 
ply houses.  The  mica  is  placed  between  the 
transistor  and  the  chassis  deck,  and  fibre  wash- 
ers are  placed  under  the  retaining  nuts  hold- 
ing the  transistor  in  place.  When  the  tran- 
sistors are  mounted  in  place,  measure  the 
collector  to  ground  resistance  with  an  ohm- 
meter.  It  should  be  100  megohms  or  higher 
in  dry  air.  After  the  mounting  is  completed, 
spray  the  transistor  and  the  bare  chassis  sec- 
tion with  plastic  Krylon  to  retard  oxidation.. 
Several  manufacturers  produce  anodized  alumi- 
num washers  that  serve  as  mounting  insula- 
tors. These  may  be  used  in  place  of  the  mica 
washers,  if  desired.  The  under-chassis  wiring 
of  the  supply  may  be  seen  in  figure  30. 

The  85  Wott  Figure  31  shows  the  schematic 
Supply  of  a dual  voltage  transistorized 

mobile  power  supply.  A bridge 


Figure  30 

UNDER-CHASSIS  LAYOUT  OF  PARTS 

Two  70  tifd.  capacitors  are  connected  in 
parallel  for  20  /xfd.  output  filter  capacitor. 
Silicon  rectifiers  are  mounted  in  dual  fuse 
clips  at  end  of  chassis.  Transistors  are  insu- 
lated from  chassis  with  thin  mica  sheets  and 
fibre  washers  if  supply  is  used  with  positive- 
grounded  primary  system. 


rectifier  permits  the  choice  of  either  250  volts 
or  500  volts,  or  a combination  of  both  at  a 
total  current  drain  that  limits  the  secondary 
power  to  85  watts.  Thus,  500  volts  at  170 
milliamperes  may  be  drawn,  with  correspond- 
ingly less  current  as  additional  power  is  drawn 
from  the  250  volt  tap. 

The  supply  is  built  upon  an  aluminum  box- 
chassis  measuring  7”x5”x3”,  the  layout  close- 
ly following  that  of  the  35  watt  supply.  Delco 
2N278  P-N-P-type  transistors  are  used  as  the 
switching  elements  and  eight  silicon  diodes 
form  the  high  voltage  bridge  rectifier.  The 
transistors  are  affixed  to  the  chassis  in  the 
same  manner  as  the  2N301A  mounting  de- 
scribed previously. 

32-8  Power  Supply 

Components 

The  usual  components  which  make  up  a 
power  supply,  in  addition  to  rectifiers  which 
have  already  been  discussed,  are  filter  ca- 
pacitors, bleeder  resistors,  transformers,  and 
chokes.  These  components  normally  will  be 
purchased  especially  for  the  intended  appli- 
cation, taking  into  consideration  the  factors 
discussed  earlier  in  this  chapter. 

Filter  There  are  two  types  of  filter  ca- 

Capoeitors  pacitors : ( 1 ) paper  dielectric 

type,  (2)  electrolytic  type. 

Paper  capacitors  consist  of  two  strips  of 
metal  foil  separated  by  several  layers  of  special 


708  Power  Supplies 


THE  RADIO 


2N278 


Figure  31 
SCHEMATIC, 

85  WATT 
TRANSISTOR 
POWER  SUPPLY 
FOR  12  VOLT 
AUTOMOTIVE 
SYSTEM 

Ti — Transistor  power 
transformer.  12  volt 
primary  to  provide  275 
volts  at  125  ma. 
Chicago  Standard 
DCr-2. 

Di-Di — Sarkes-Tarzian 
silicon  rectifier,  type 
M-SOO 


paper.  Some  types  of  paper  capacitors  are 
wax-impregnated,  but  the  better  ones,  especial- 
ly the  high-voltage  types,  are  oil-impregnated 
and  oil-filled.  Some  capacitors  are  rated  both 
for  flash  test  and  normal  operating  voltages; 
the  latter  is  the  important  rating  and  is  the 
maximum  voltage  which  the  capacitor  should 
be  required  to  withstand  in  service. 

The  capacitor  across  the  rectifier  circuit  in 
a capacitor-input  filter  should  have  a ivo  'king 
voltage  rating  equal  at  least  to  1.41  times  the 
r-m-s  voltage  output  of  the  reaifier.  The  re- 
maining capacitors  may  be  rated  mote  nearly 
in  accordance  with  the  d-c  voltage. 

The  electrolytic  capacitor  consists  of  two 
aluminum  electrodes  in  contact  with  a conduct- 
ing paste  or  liquid  which  acts  as  an  electro- 
lyte. A very  thin  film  of  oxide  is  formed  on  the 
surface  of  one  electrode,  called  the  anode. 
This  film  of  oxide  acts  as  the  dielectric.  The 
electrolytic  capacitor  must  be  correctly  con- 
nected in  the  circuit  so  that  the  anode  always 
is  at  a positive  potential  with  respect  to  the 
electrolyte,  the  latter  actually  serving  as  the 
other  electrode  (plate)  of  the  capacitor.  A re- 
versal of  the  polarity  for  any  length  of  time 
will  ruin  the  capacitor. 

The  dry  type  of  electrolytic  capacitor  uses 
an  elearolyte  in  the  form  of  paste.  The  die- 
lectric in  electrolytic  capacitors  is  not  perfect; 
these  capacitors  have  a much  higher  direct 
current  leakage  than  the  paper  type. 

The  high  capacitance  of  electrolytic  ca- 
pacitors results  from  the  thinness  of  the  film 
which  is  formed  on  the  plates.  The  maximum 
voltage  that  can  be  safely  impressed  across 
the  average  electrolytic  filter  capacitor  is  be- 
tween 450  and  600  volts;  the  working  voltage 
is  usually  rated  at  450.  When  electrolytic  ca- 


pacitors are  used  in  filter  circuits  of  high- 
voltage  supplies,  the  capacitors  should  be  con- 
nected in  series.  The  positive  terminal  of  one 
capacitor  must  connect  to  the  negative  ter- 
minal of  the  other,  in  the  same  manner  as 
dry  batteries  are  connected  in  series. 

It  is  not  necessary  to  connect  shunt  resis- 
tors across  each  electrolytic  capacitor  section 
as  it  is  with  paper  capacitors  connected  in 
series,  because  electrolytic  capacitors  have 
fairly  low  internal  d-c  resistance  as  compared 
to  paper  capacitors.  Also,  if  there  is  any  varia- 
tion in  resistance,  it  is  that  electrolytic  unit 
in  the  poorest  condition  which  will  have  the 
highest  leakage  current,  and  therefore  the  volt- 
age across  this  capacitor  will  be  lower  than 
that  across  one  of  the  series  connected  units 
in  better  condition  and  having  higher  internal 
resistance.  Thus  we  see  that  equalizing  re- 
sistors are  not  only  unnecessary  across  series- 
connected  electrolytic  capacitors  but  are  ac- 
tually undesirable.  This  assumes,  of  course, 
similar  capacitors  by  the  same  manufacturer 
and  of  the  same  capacitance  and  voltage  rat- 
ing. It  is  not  advisable  to  connect  in  series 
electrolytic  capacitors  of  different  make  or 
ratings. 

There  is  very  little  economy  in  using  elec- 
trolytic capacitors  in  series  in  circuits  where 
more  than  two  of  these  capacitors  would  be 
required  to  prevent  voltage  breakdown. 

Electrolytic  capacitors  can  be  greatly  re- 
duced in  size  by  the  use  of  etched  aluminum 
foil  for  the  anode.  This  greatly  increases  the 
surface  area,  and  the  dielectric  film  covering 
it,  but  raises  the  power  factor  slightly.  For 
this  reason,  ultra-midget  electrolytic  capacitors 
ordinarily  should  not  be  used  at  full  rated  d-c 
voltage  when  a high  a-c  component  is  present. 


HANDBOOK 


Special  Supplies  709 


such  as  would  be  the  case  for  the  input  ca- 
pacitor in  a capacitor-input  filter. 

Bleeder  A heavy  duty  resistor  should  be 
Resistors  connected  across  the  output  of  a 
filter  in  order  to  draw  some  load 
current  at  all  time.  This  resistor  avoids  soar- 
ing of  the  voltage  at  no  load  when  swinging 
choke  input  is  used,  and  also  provides  a means 
for  discharging  the  filter  capacitors  when  no 
external  vacuum-tube  circuit  load  is  connected 
to  the  filter.  This  bleeder  resistor  should  nor- 
mally draw  approximately  10  per  cent  of  the 
full  load  current. 

The  power  dissipated  in  the  bleeder  resistor 
can  be  calculated  by  dividing  the  square  of 
the  d-c  voltage  by  the  resistance.  This  power 
is  dissipated  in  the  form  of  heat,  and,  if  the 
resistor  is  not  in  a well-ventilated  position, 
the  wattage  rating  should  be  higher  than  the 
actual  wattage  being  dissipated.  High-voltage, 
high-capacitance  filter  capacitors  can  hold  a 
dangerous  charge  if  not  bled  off,  and  wire- 
wound  resistors  occasionally  open  up  without 
warning.  Hence  it  is  wise  to  place  carbon  re- 
sistors in  series  across  the  regular  wire-wound 
bleeder. 

When  purchasing  a bleeder  resistor,  be  sure 
that  the  resistor  will  stand  not  only  the  re- 
quited wattage,  but  also  the  voltage.  Some  re- 
sistors have  a voltage  limitation  which  makes 
it  impossible  to  force  sufficient  current  through 
them  to  result  in  rated  wattage  dissipation. 
This  type  of  resistor  usually  is  provided  with 
slider  taps,  and  is  designed  for  voltage  divider 
service.  An  untapped,  non-adjustable  resistor 
is  preferable  as  a high  voltage  bleeder,  and  is 
less  expensive.  Several  small  resistors  may  be 
connected  in  series,  if  desired,  to  obtain  the 
required  wattage  and  voltage  rating. 

Transformers  Power  transformers  and  fila- 
ment transformers  normally 
will  give  no  trouble  over  a period  of  many 
years  if  purchased  from  a reputable  manufac- 
turer, and  if  given  a reasonable  amount  of 
care.  Transformers  must  be  kept  dry;  even  a 
small  amount  of  moisture  in  a high-voltage 
unit  will  cause  quick  failure.  A transformer 
which  is  operated  continuously,  within  its 
ratings,  seldom  will  give  trouble  from  mois- 
ture, since  an  economically  designed  trans- 
former operates  at  a moderate  temperamre 
rise  above  the  temperature  of  the  surrounding 
air.  But  an  unsealed  transformer  which  is 
inactive  for  an  appreciable  period  of  time  in 
a highly  humid  location  can  absorb  enough 
moisture  to  cause  early  failure. 


Pilfer  Choke  Filter  inductors  consist  of  a 
Coils  coil  of  wire  wound  on  a lami- 

nated iron  core.  The  size  of 
wire  is  determined  by  the  amount  of  direa 
current  which  is  to  flow  through  the  choke 
coil.  This  direct  current  magnetizes  the  core 
and  reduces  the  inductance  of  the  choke  coil; 
therefore,  filter  choke  coils  of  the  smoothing 
wpe  ate  built  with  an  air  gap  of  a small  frac- 
tion of  an  inch  in  the  iron  core,  for  the  pur- 
pose of  preventing  saturation  when  maximum 
d.c.  flows  through  the  coil  winding.  The  "air 
gap”  is  usually  in  the  form  of  a piece  of  fiber 
inserted  between  the  ends  of  the  laminations. 
The  air  gap  reduces  the  initial  inductance  of 
the  choke  coil,  but  keeps  it  at  a higher  value 
under  maximum  load  conditions.  The  coil  must 
have  a great  many  more  turns  for  the  same 
initial  inductance  when  an  air  gap  is  used. 

The  d-c  resistance  of  any  filter  choke  should 
be  as  low  as  practicable  for  a specified  value 
of  inductance.  Smaller  filter  chokes,  such  as 
those  used  in  radio  receivers,  usually  have 
an  inductance  of  from  6 to  15  henrys,  and  a 
d-c  resistance  of  from  200  to  400  ohms.  A 
high  d-c  resistance  will  reduce  the  output  volt- 
age, due  to  the  voltage  drop  across  each  choke 
coil.  Large  filter  choke  coils  for  radio  trans- 
mitters and  Class  B amplifiers  usually  have 
less  than  100  ohms  d-c  resistance. 

32-9  Special  Power 
Supplies 

A complete  transmitter  usually  includes  one 
or  more  power  supplies  such  as  grid-bias 
packs,  voltage-regulated  supplies,  or  trans- 
formerless supplies  having  some  special  char- 
acteristic. 

Regulated  Where  it  is  desired  in  a circuit 
Supplies — to  stabilize  the  voltage  supply 
V-R  T ubes  to  a load  requiring  not  mote  than 
perhaps  20  to  25  ma,,  the  glow- 
discharge  type  of  voltage-regulator  tube  can 
be  used  to  great  advantage.  Examples  of  such 
circuits  are  the  local  oscillator  circuit  in  a 
receiver,  the  tuned  oscillator  in  a v-f-o-,  the 
oscillator  in  a frequency  meter,  or  the  bridge 
circuit  in  a vacuum-tube  voltmeter.  A num- 
ber of  tubes  are  available  for  this  applica- 
tion including  the  OA3/VR75,  OB3/’VR90, 
OC3/VR105,  OD3/VR150,  and  the  OA2 
and  OB2  miniature  types.  These  tubes  stabil- 
ize the  voltage  across  their  terminals  to  75, 
90,  105,  or  130  volts.  The  minature  types 
OA2  stabilize  to  150  volts  and  OB2  to  108 
volts.  The  types  OA2,  OB2  and  OB3/VR90 


710  Power  Supplies 


THE  RADIO 


have  a maximum  current  rating  of  30  ma. 

and  the  other  three  types  have  a maximum 
current  rating  of  40  ma.  The  minimum  cur- 
rent required  by  all  six  types  to  sustain  a 
constant  discharge  is  5 ma. 

A VR  tube  (common  term  applied  to  all 
glow-discharge  voltage  regulator  tubes)  may 
be  used  to  stabilize  the  voltage  across  a vari- 
able load  or  the  voltage  across  a constant  load 
fed  from  a varying  voltage.  Two  or  more  VR 
tubes  may  be  connected  in  series  to  provide 
exactly  180,  210,  255  volts  or  other  combi- 
nations of  the  voltage  ratings  of  the  tubes. 
It  is  not  recommended,  however,  that  VR 
tubes  be  connected  in  parallel  since  both  the 
striking  and  the  regulated  voltage  of  the 
paralleled  tubes  normally  will  be  sufficiently 
different  so  that  only  one  of  the  tubes  will 
light.  The  remarks  following  apply  generally 
to  all  the  VR  types  although  some  examples 
apply  specifically  to  the  OD3/VR150  type. 

A device  requiring  say,  only  50  volts  can 
be  stabilized  against  supply  voltage  variations 
by  means  of  a VR-105  simply  by  putting  a 
suitable  resistor  in  series  with  the  regulated 
voltage  and  the  load,  dropping  the  voltage 
from  105  to  50  volts.  However,  it  should  be 
borne  in  mind  that  under  these  conditions 
the  device  will  not  be  regulated  for  varying 
load;  in  other  words,  if  the  load  resistance 
varies,  the  voltage  across  the  load  will  vary, 
even  though  the  regulated  voltage  remains  at 
105  volts. 

To  maintain  constant  voltage  across  a vary- 
ing load  resistance  there  must  be  no  series 
resistance  between  the  regulator  tube  and  the 
load.  This  means  that  the  device  must  be 
operated  exactly  at  one  of  the  voltages  ob- 
tainable by  seriesing  two  or  more  similar  or 
different  VR  tubes. 

A VR-150  may  be  considered  as  a stubborn 
variable  resistor  having  a range  of  from  30,000 
to  5000  ohms  and  so  intent  upon  maintaining 
a fixed  voltage  of  150  volts  across  its  ter- 
minals that  when  connected  across  a voltage 
source  having  very  poor  regulation  it  will  in- 
stantly vary  its  own  resistance  within  the  lim- 
its of  5000  and  30,000  ohms  in  an  attempt 
to  maintain  the  same  150  volt  drop  across  its 
terminals  when  the  supply  voltage  is  varied. 
The  theory  upon  which  a VR  tube  operates  is 
covered  in  an  earlier  chapter. 

It  is  paradoxical  that  in  order  to  do  a good 
job  of  regulating,  the  regulator  tube  must  be 
fed  from  a voltage  source  having  poor  regula- 
tion (high  series  resistance).  The  reason  for 
this  presently  will  become  apparent. 


If  a high  resistance  is  connected  across  the 
VR  tube,  it  will  not  impair  its  ability  to  main- 
tain a fixed  voltage  drop.  However,  if  the  load 
is  made  too  low,  a variable  5000  to  30,000 
ohm  shunt  resistance  (the  VR-150)  will  not 
exert  sufficient  effect  upon  the  resulting  re- 
sistance to  provide  constant  voltage  except 
over  a very  limited  change  in  supply  voltage 
or  load  resistance.  The  tube  will  supply  maxi- 
mum regulation,  or  regulate  the  largest  load, 
when  the  source  of  supply  voltage  has  high 
internal  or  high  series  resistance,  because  a 
variation  in  the  effective  internal  resistance 
of  the  VR  tube  will  then  have  more  control- 
ling effect  upon  the  load  shunted  across  it. 

In  order  to  provide  greatest  range  of  regu- 
lation, a VR  tube  (or  two  in  series)  should  be 
used  with  a series  resistor  (to  effect  a poorly 
regulated  voltage  source)  of  such  a value  that 
it  will  permit  the  VR  tube  to  draw  from  8 to 
20  ma.  under  normal  or  average  conditions 
of  supply  voltage  and  load  impedance.  For 
maximum  control  range,  the  series  resistance 
should  be  not  less  than  approximately  20,000 
ohms,  which  will  necessitate  a source  of  volt- 
age considerably  in  excess  of  150  volts.  How- 
ever, where  the  supply  voltage  is  limited,  good 
control  over  a limited  range  can  be  obtained 
with  as  little  as  3000  ohms  series  resistance. 
If  it  takes  less  than  3000  ohms  series  resist- 
ance to  make  the  VR  tube  draw  15  to  20 
ma.  when  the  VR  tube  is  connected  to  the 
load,  then  the  supply  voltage  is  not  high 
enough  for  proper  operation. 

Should  the  current  through  a VR-150,  VR- 
105,  or  VR-75  be  allowed  to  exceed  40  ma., 
the  life  of  the  tube  will  be  shortened.  If  the 
current  falls  below  5 ma.,  operation  will  be- 
come unstable.  Therefore,  the  tube  must  op- 
erate within  this  range,  and  within  the  two 
extremes  will  maintain  the  voltage  within  1.5 
per  cent.  It  takes  a voltage  excess  of  at  least 
10  to  15  per  cent  to  "start”  a VR  type  regu- 
lator; and  to  insure  positive  starting  each  time 
the  voltage  supply  should  preferably  exceed 
the  regulated  output  voltage  rating  about  20 
per  cent  or  more.  This  usually  is  automatically 
taken  care  of  by  the  fact  that  if  sufficient 
series  resistance  for  good  regulation  is  em- 
ployed, the  voltage  impressed  across  the  VR 
tube  before  the  VR  tube  ionizes  and  starts 
passing  current  is  quite  a bit  higher  than  the 
starting  voltage  of  the  tube. 

When  a VR  tube  is  to  be  used  to  regulate 
the  voltage  applied  to  a circuit  drawing  less 
than  15  ma.  normal  or  average  current,  the 
simplest  method  of  adjusting  the  series  resist- 


HANDBOOK 


Special  Supplies  711 


RS,  SERIES  RESISTOR 
O Wr 

Es 

D,C.  SUPPLY 
VOLTAGE 

O 

Figure  32 

STANDARD  VR-TUBE  REGULATOR 
CIRCUIT 

The  VR-tube  regulator  will  maintain  the  volt< 
age  across  its  terminals  constant  within  a 
few  volts  for  moderate  voriafions  in  or 
See  text  for  discussion  of  the  use  of  VR 
tubes  in  various  circuit  applications. 


ance  is  to  remove  the  load  and  vary  the  series 
resistor  until  the  VR  tube  draws  about  30  ma. 
Then  connect  the  load,  and  that  is  all  there  is 
to  it.  This  method  is  particularly  recommended 
when  the  load  is  a heater  type  vacuum  tube, 
which  may  not  draw  current  for  several  sec- 
onds after  the  power  supply  is  mrned  on. 
Under  these  conditions,  the  current  through 
the  VR  tube  will  never  exceed  40  ma.  even 
when  it  is  running  unloaded  (while  the  heater 
tube  is  warming  up  and  the  power  supply 
rectifier  has  already  reached  operating  tem- 
perature ) . 

Figure  32  illustrates  the  standard  glow  dis- 
charge regulator  tube  circuit.  The  mbe  will 
maintain  the  voltage  across  Rl  constant  to 
within  1 or  2 volts  for  moderate  variations  in 
Rl  or  Es. 

Voltage  Regulated  When  it  is  desired  to  sta- 
Power  Supplies  bilize  the  potential  across 
a circuit  drawing  more 
than  a few  milliamperes  it  is  advisable  to 
use  a voltage-regulated  power  supply  of  the 
type  illustrated  in  figure  33  rather  than  glow 
discharge  tubes. 

A 6AS7-G  is  employed  as  the  series  control 
element,  and  type  816  mercury  vapor  reaifiers 
are  used  in  the  power  supply  section.  The 
6AS7-G  acts  as  a variable  series  resistance 
which  is  controlled  by  a separate  regulator 
tube  much  in  the  manner  of  a-v-c  circuits  or 
inverse  feedback  as  used  in  receivers  and  a-f 
amplifiers.  A 6SH7  controls  the  operating  bias 
on  the  6AS7-G,  and  therefore  controls  the  in- 
ternal resistance  of  the  6AS7-G.  This,  in  turn, 
controls  the  output  voltage  of  the  supply,  which 
controls  the  plate  current  of  the  6SH7,  thus 
completing  the  cycle  of  regulation.  It  is  ap- 
parent that  under  these  conditions  any  change 
in  the  output  voltage  will  tend  to  "resist  it- 
self,” much  as  the  a-v-c  system  of  a receiver 
resists  any  change  in  signal  strength  delivered 
to  the  detector. 


Because  it  is  necessary  that  there  always  be 
a moderate  voltage  drop  through  the  6AS7-G 
in  order  for  it  to  have  proper  control,  the  rest 
of  the  power  supply  is  designed  to  deliver  as 
much  output  voltage  as  possible.  This  calls 
for  a low  resistance  full-wave  rectifier,  a high 
capacitance  output  capacitor  in  the  filter  sys- 
tem and  a low  resistance  choke. 

Reference  voltage  in  the  power  supply  is 
obtained  from  a VR-150  gaseous  regulator. 
Note  that  the  6.3 -volt  heater  winding  for  the 
6SH7  and  the  6AS7-G  tubes  is  operated  at 
a potential  of  plus  150  volts  by  conneaing 
the  winding  to  the  plate  of  the  VR-150.  This 
procedure  causes  the  heater-cathode  voltage 
of  the  6SH7  to  be  zero,  and  permits  an  out- 
put voltage  of  up  to  450  since  the  300-volt 
heater-to-cathode  rating  of  the  6AS7-G  is  not 
exceeded  with  an  output  voltage  of  450  from 
the  power  supply. 

The  6SH7  tube  was  used  in  place  of  the 
more  standard  6SJ7  after  it  was  found  that 
the  regulation  of  the  power  supply  could  be 
improved  by  a factor  of  two  with  the  6SH7 
in  place  of  the  6SJ7.  The  original  version  of 
the  power  supply  used  a 5R4-GY  rectifier 
tube  in  place  of  the  8l6’s  which  now  are 
used.  The  excessive  drop  of  the  5R4-GY  re- 
sulted in  loss  of  control  by  the  regulator  por- 


REGULATED  POWER  SUPPLY 
Ti — 615  or  520  volts  each  side  of  c.t.,  300 

ma.  Stancof  P-8041. 

Ti — 5 volts  at  3 amp.,  6.3  volts  at  6 amp. 
Stancor  P-5009. 

CH — 4 henry  at  250  ma.  Stancor  C-1412. 


712  Power  Supplies 


THE  RADIO 


Ti  Vi 


COMPONENTS 

APPROXIMATE 
OUTPUT  VOLTAGE 

MAX. 

CURRENT 

6.3  V. 
FILAMENT 

Ti 

Vi 

CH  1 

CH2 

Cl 

C2 

Rl 

NO 

LOAD 

FULL 

LOAD 

350-0-350 
SrAA/COff  PC-B409 
MER/r  R-31SZ 

5Y3-GT 

10  H. 

STANCOR 

C-1Q01 

MERIT 

C-Z99% 

10H. 

STANCOR 

C-fOOt 

MERIT 

C-2993 

lOUF.  450  V. 
CORNELL- 
OUBILIER 
BR-t04S 

20JJF,450V. 
CORN  ELL- 
DUB!  LIE  R 
BR-204S 

35K.10W 

310 

240 

60  MA. 

3 A. 

37S-0-37S 
STANCOR  PC-OAH 
MERIT  P-29S4 

5Y3-GT 

3-13  H 

STANCOR 

C-1716 

MERIT 

C-3167 

1 H. 

STANCOR 

C~1421 

MERIT 

C-3160 

10JUF.450V. 

CORNELL- 

DUBILIER 

BR-104S 

20  JUF,  450  V. 
CORNELL- 
DUBILIER 
BR-Z04S 

35K,10W 

330 

230 

1 

140  MA. 

4.5A. 

400-0-400 
STANCOR  PC-6413 

5U4-G 

2-12  H. 

STANCOR 

C-140Z 

MERIT 

C-3169 

4 H. 

STANCOR 

C-1412 

MERIT 

C-3I6Z 

tOUF  450  V. 
COR  N ELL- 
DUB!  LIE  R 
BR-1046 

10  UF. 450  V. 
CORNELL- 
DUBILIER 

8 R- 1043 

35K.10W 

360 

270 

250  MA. 

SA. 

525-0-325 
UTC  S-40 

5U4-GB 

5-25  H. 
UTC  S-32 

20  H. 

UTC  S-3f 

1OJUF,600V. 

MALLORY 

TC-92 

101iP,600V. 

MALLORY 

TC-9Z 

35K.10W 

460 

375 

240  MA. 

4 A. 

600-0-600 
tITC  S-41 

5R4-GY 

5-25  H. 

UTC  S~32 

20  H. 

UTC  s~3r 

6JUF.  600  V. 
SPRA6UE 
CR~66 

6iiF,600  V. 
SPRA6UE 
CR-66 

35  K,25W 

540 

410 

200  MA. 

4A. 

900-0-900 
UTC  S-45 

5R4-aY 

3-25  H. 

UTC  S-3E 

20  H. 

UTC  S~3f 

4UF,  1 KV. 
SPRABUE 
CR-41 

eur.  1KV. 
SPRABUE 
CR-61 

50K,25W, 

630 

650 

t75MA. 

- 

Figure  34 

DESIGN  CHART  FOR  CHOKE-INPUT  POWER  SUPPLIES 


tion  with  an  output  voltage  of  about  390 
with  a 225-ma.  drain.  Satisfactory  regulation 
can  be  obtained,  however,  at  up  to  450  volts 
if  the  maximum  current  drain  is  limited  to 
150  ma.  when  using  a 5R4-GY  rectifier.  If 
the  power  transformer  is  used  with  the  taps 
giving  520  volts  each  side  of  center,  and  if 
the  maximum  drain  is  limited  to  225  ma., 
a type  83  rectifier  may  be  used  as  the  power 
supply  rectifier.  The  615-volt  taps  on  the 
power  transformer  deliver  a voltage  in  excess 
of  the  maximum  ratings  of  the  83  tube.  With 
the  83  in  the  power  supply,  excellent  regu- 
lation may  be  obtained  with  up  to  about  420 
volts  output  if  the  output  current  is  limited 
to  225  ma.  But  with  the  816’s  as  rectifiers 
the  full  capabilities  of  all  the  components 
in  the  power  supply  may  be  utilized. 

If  the  power  supply  is  to  be  used  with  an 
output  voltage  of  400  to  450  volts,  the  full 
615  volts  each  side  of  center  should  be  ap- 
plied to  the  8l6’s.  However,  the  maximum 
plate  dissipation  rating  of  the  6AS7-G  will 
be  exceeded,  due  to  the  voltage  drop  across 
the  tube,  if  the  full  current  rating  of  250  ma. 
is  used  with  an  output  voltage  below  400 
volts.  If  the  power  supply  is  to  be  used  with 
full  output  current  at  voltages  below  400  volts 
the  520-volt  taps  on  the  plate  transformer 


should  be  connected  to  the  8l6’s.  Some  varia- 
tion in  the  output  range  of  the  power  supply 
may  be  obtained  by  varying  the  values  of  the 
resistors  and  the  potentiometer  across  the  out- 
put. However,  be  sure  that  the  total  plate 
dissipation  rating  of  26  watts  on  the  6AS7-G 
series  regulator  is  not  exceeded  at  maximum 
current  output  from  the  supply.  The  total  dis- 
sipation in  the  6AS7-G  is  equal  to  the  cur- 
rent through  it  (output  current  plus  the  cur- 
rent passing  through  the  two  bleeder  strings) 
multiplied  by  the  drop  through  the  tube  (volt- 
age across  the  filter  capacitor  minus  the  out- 
put voltage  of  the  supply ) . 

32-10  Power  Supply 
Design 

Power  supplies  may  either  be  of  the  choke 
input  type  illustrated  in  figure  34,  or  the  ca- 
pacitor input  type,  illustrated  in  figure  35.  Ca- 
pacitor input  filter  systems  are  characterized 
by  a d-c  supply  output  voltage  that  runs  from 
0.9  to  about  1.3  times  the  r.m.s.  voltage  of 
one-half  of  the  high  voltage  secondary  wind- 
ing of  the  transformer.  The  approximate  regu- 
lation of  a capacitor  input  filter  system  is 
shown  in  figure  36.  Capacitor  input  filter 
systems  are  not  recommended  for  use  with 


HANDBOOK 


Special  Supplies  713 


Tl  Vi 


COMPONENTS 

APPROXIMATE 
OUTPUT  VOLTAGE 

MAX. 

CURRENT 

6.3V. 

FILAMENT 

Ti 

Vl 

CHl 

C2 

Rl 

NO 

LOAD 

FULL 

LOAD 

290-0-260 
STANCOR  PC-a404 
MERIT  P-3 14a 

5Y3-GT 

10  H. 

STANCOR 

C-IOOf 

MERIT 

C-Z993 

20iJF,4S0V. 

CORNEU- 

DUBILIER 

aR-Z04S 

20  JJF,  4S0  V. 
CORNELL- 
DUBILIER 
BR-Z04S 

3SK,10W 

340 

240 

60  MA. 

3A. 

375-0-37S 
STANCOR  PC-a4U 
MERIT  P-E954 

— 

SY3-GT 

7 H. 

STANCOR 

C-t42l 

MERIT 

c-3iao 

lOiJF.eOOV. 

MALLORY 

TC-9Z 

10JJF,e00  V. 
MALLORY 
TC-9Z 

35  K, tow 

460 

350 

125  MA. 

4.5A. 

435-0-43S 
MERIT  P-3t56 

5U4-G 

4 H. 

STANCOR 

C-14i2 

MERIT 

c-siaz 

SJJF.OOOV. 

SPRAOUE 

cR-aa 

6 UF,  600  V. 
SPRAQUE 
cR-aa 

35K,25W 

600 

400 

225  MA. 

6 A. 

600-0-600 
STANCOR  PC-a414 

SR4-GY 

4 H. 

STANCOR 

C-I4IZ 

MERIT 

c-3  laz 

4 JJF,  1 KV. 
SPRAEUE 
CR-41 

6 UF,  KV. 
SPRAEUE 
CR-B! 

50K,25W 

600 

600 

200  MA. 

OA. 

900-0-900 
UTC  S-45 

5R4-GY 

20  H. 
t/TC  S~3/ 

4UF»  1.SKV. 

SPRAEUE 

CR-41S 

6UF.  1.6KV. 

SPRAEUE 

cR-an 

75K  25W 

1200 

910 

150  MA. 

- 

Figure  35 

DESIGN  CHART  FOR  CAPACITOR-INPUT  POWER  SUPPLIES 


mercury  vapor  rectifier  cubes,  as  the  peak  rec- 
tifier current  may  run  as  high  as  five  or  six 
times  the  d-c  load  current  of  the  power  sup- 
ply. It  is  possible,  however,  to  employ  type 
872-A  mercury  vapor  rectifier  tubes  in  ca- 
pacitor input  circuits  wherein  the  load  cur- 
rent is  less  than  600  milliamperes  or  so,  and 
where  a low  resistance  bleeder  is  used  to  hold 
the  minimum  current  drain  of  the  supply  to 
a value  greater  than  50  milliamperes  or  so. 
Under  these  conditions  the  peak  plate  current 
of  the  872-A  mercury  vapor  tubes  will  not 
be  exceeded  if  the  input  filter  capacitor  is 
4 /rfd.  or  less. 

Choke  input  filter  systems  ate  characterized 
by  lower  peak  load  currents  (1.1  to  1.3  times 
the  average  load  current)  than  the  capacitor 
input  filter,  and  by  better  voltage  regulation. 
Design  Charts  for  capacitor  and  choke  input 
filter  supplies  for  various  voltages  and  load 
currents  are  shown  in  figures  34,  35,  and  37. 

The  construction  of  power  supplies  for  trans- 
mitters, receivers  and  accessory  equipment  is 
a relatively  simple  matter  electrically  since 
lead  lengths  and  placement  of  parts  are  of 
minor  importance  and  since  the  circuits  them- 
selves are  quite  simple.  Under-chassis  wiring 
of  a heavy-duty  supply  is  shown  in  figure  38. 

Bridge  Supplies  Some  practical  variations  of 
the  common  bridge  rectifier 


circuit  of  figure  6 are  illustrated  in  figures  39 
and  40.  In  many  instances  a transmitter  or 
modulator  requites  two  different  supply  volt- 
ages, differing  by  a ratio  of  about  2:1.  A 
simple  bridge  supply  such  as  shown  in  figure 
39  will  provide  both  of  these  voltages  from 
a simple  broadcast  "replacement-type”  power 
transformer.  The  first  supply  of  figure  39  is 
ample  to  power  a transmitter  of  the  6CL6-807 
type  to  an  input  of  60  watts.  The  second  sup- 
ply will  run  a transmitter  running  up  to  120 
watts,  such  as  one  employing  a pair  of  6146 


RATIO  OF  D.C.  OUTPUT  VOLTAGE  TO  R.M.S  VOLTAGE 
OF  SECONDARY  WINDING  OF  PLATE  TRANSFORMER. 

Figure  36 

APPROXIMATE  REGULATION  OF 
CAPACITOR-INPUT  FILTER  SYSTEM 


714  Power  Supplies 


THE  RADIO 


T2 


COMPONENTS 

APPROXIMATE 
OUTPUT  VOLTAGE 

MAX. 

CURRENT 

(ICAS) 

Ti 

T2 

Vi- V2 

CHi 

CH2 

Cl 

CZ 

Rl 

NO 

LOAD 

FULL 

LOAD 

1150-0-1150 

C»/CAGOrfiA/VS. 

P-J07 

2.5  V..  10  A. 
CHI  TRAN. 
F-210 

86e-A 

aee-A 

SH. 

CHI.  TRAN. 
R-65 

lOH. 

CHI.  TRAN. 
R-105 

4UF.  1.5KV. 
SAN6AMO 
7115-4 

tUF,  1.SKV 
SANGAMO 
7115-6 

40K,75W 

1150 

1000 

350  MA. 

1710-0-1710 
CHICAGO  TRANS. 
P-1512 

2.5V., 10A. 
CHI.  TRAN. 
F-210H 

aee-A 

86e-A 

8 H. 

CHI.  TRAN. 
R-65 

10H. 

CHI.  TRAN. 
R-105 

4UF,  2 KV. 
SANGAMO 
7120-4 

8XIF,  2KV. 
SANGAMO 
7120-6 

50  K.75W 

1700 

1500 

425  MA. 

2900-0-2900 
CHICAGO  TRANS. 
P-2126 

5 V.,  10A. 
CHI.  TRAN. 
F-51QH 

872-A 

872-A 

6 H. 

CHI.  TRAN. 
R-67 

6 H. 

CHI.  TRAN. 
R-67 

4IIF,  3KV. 
SANGAMO 
7150-4 

4iiF,  3KV. 
SANGAMO 
7150-4 

75K 
200  W 

2750 

2500 

700  MA. 

3500-0-3500 
UTC  CG-509 

5 V.,  10A. 
UTC 
LS-62 

872-A 

872-A 

10  H. 
UTC 
CG-1S 

10  H. 
UTC 
CG-1S 

4JJF.  4KV. 
CORNELL- 
OU OILIER 

T 40040- A 

4JUF,  4KV. 
CORN ELL- 
DUB!  HER 
T40040-A 

100  K 
200W 

3400 

3000 

1000  MA. 

4600-0-4600 
UTC  CG-510 
CHICAGO  TRANS. 
P-4555 

5V-,20A. 

UTC 

LS-65 

575-A 

3 75-A 

10  H. 
UTC 
CG-tS 

10  H. 

UTC 

CG-tS 

4JJF.  5KV. 
AERO  VOX 
Jfi-09 

4XJF,  5KV. 
AEROVOX 
JP-09 

100  K 
300  W 

4400 

4000 

800  MA. 

Figure  37 

DESIGN  CHART  FOR  CHOKE-INPUT  HIGH  VOLTAGE  SUPPLIES 


tetrodes  in  the  power  amplifier  stage.  It  is 
to  be  noted  that  separate  filament  transform- 
ers are  used  for  rectifier  tubes  Vi  and  V2,  and 
that  one  leg  of  each  filament  is  connected  to 
the  cathode  of  the  respective  tube,  which  is 
at  a high  potential  with  respect  to  ground. 
The  choke  CHi  in  the  negative  lead  of  the 
supply  serves  as  a common  filter  choke  for 
both  output  voltages.  Each  portion  of  the  sup- 
ply may  be  considered  as  having  a choke  in- 
put filter  system.  Filaments  of  Vi  should  be 
energized  before  the  primary  voltage  is  ap- 
plied to  Ti. 


Bridge  supplies  may  also  be  used  to  ad- 
vantage to  obtain  relatively  high  plate  volt- 
ages for  high  powered  transmitting  equip- 
ment. Type  866-A  and  872-A  rectifier  tubes 
can  only  serve  in  a supply  delivering  under 
3500  volts  in  a full-wave  circuit.  Above  this 
voltage,  the  peak  inverse  voltage  rating  of  the 
rectifier  tube  will  be  exceeded,  and  danger 
of  flash-back  within  the  rectifier  tube  will  be 
present.  However,  with  bridge  circuits,  the 
same  tubes  may  deliver  up  to  as  much  as  7000 
volts  d.c.  without  exceeding  the  peak  inverse 
voltage  rating. 

The  bridge  circuit  also  permits  the  use  of 
the  so-called  "pole  transformer’’  in  high  volt- 
age power  supplies.  Two  KVA  transformers  of 
this  type  having  a 110/220  volt  secondary 
winding  and  a split  2200  volt  primary  wind- 
ing may  often  be  picked  up  in  salvage  yards 
for  a dollar  or  two.  If  reversed,  and  either 
110  or  220  volts  applied  to  the  "primary” 
winding  approximately  2200  volts  r.m.s.  will 
be  developed  across  the  new  "secondary”  wind- 
ing. If  used  in  a bridge  circuit  as  shown  in 


Figure  38 

UNDER-CHASSIS 
POWER  SUPPLY  ASSEMBLY 

All  components  are  firmly  mounted  to  the 
steel  chassis  and  all  wiring  is  cabled.  High 
voltage  leads  are  run  in  automobile  ignition 
cables.  Heavy-duty  terminal  strips  are  mounted 
along  the  rear  edge  of  the  chassis.  The  con- 
trol panel  of  this  supply  is  shown  in  figure  7 
of  this  chapter. 


HANDBOOK 


Special  Supplies  715 


1 


F 


O- 


Ta 

F 


T2,T3  = 6.3V.,  1 A, 
STANCOR  P-6n4 


COMPONENTS 

FULL  LOAD  VOLT. 

MAX,  CURRENT 

Ti 

V1-V2 

V3 

C1-C2 

C3-C4 

CHl 

CH2 

R1-R2 

Rs 

HV»1 

HV4r2 

# 1 

«2 

360-0-360 
STANCOR  PC-B410 

6X5-GT 

5V4-C 

I6JJF 
4S0  V. 

16JJF. 
450  V. 

10H. 
120  MA. 

8 H. 

50MA. 

20K,10W. 

100  K 

1 W 

600 

240 

100  MA. 

40MA. 

400-0-400 
STANCOR  PC-A41Z 

6AX5-GT 

SU4-GB 

I6JUF. 
450  V. 

ISAJF. 

450  V. 

10  H. 
225  MA. 

8 H. 
75MA. 

20  K 10  W. 

100  K 

1 V 

625 

260 

175  MA. 

50  MA. 

Figure  39 

DUAL  VOLTAGE  BRIDGE  POWER  SUPPLIES 


figure  40,  a d-c  supply  voltage  of  about  1900 
volts  at  a current  of  500  milliamperes  may  be 
drawn  from  such  a transformer.  Do  not  at- 
tempt to  use  a smaller  transformer  than  the 
2-KVA  rating,  as  the  voltage  regulation  of 
the  unit  will  be  too  poor  for  practical  pur- 
poses. 

For  higher  voltages,  a pole  transformer  with 
a 4400  volt  primary  and  a 110/220  volt  sec- 
ondary may  be  reversed  to  provide  a d-c  plate 
supply  of  about  3800  volts. 

Commercial  plate  transformers  intended  for 
full  wave  rectifier  service  may  also  be  used 
in  bridge  service  provided  that  the  insulation 
at  the  center-tap  point  of  the  high  voltage 
winding  is  sufficient  to  withstand  one-half  of 
the  r.m.s.  voltage  of  the  secondary  winding. 
Many  high  voltage  transformers  are  specifical- 
ly designed  for  operation  with  the  center-tap 
of  the  secondary  winding  at  ground  potential; 
consequently  the  insulation  of  the  winding  at 
this  point  is  not  designed  to  withstand  high 
voltage.  It  is  best  to  check  with  the  manufac- 
turer of  the  transformer  and  find  out  if  the 
insulation  will  withstand  the  increased  volt- 


age before  a full  wave-type  transformer  is 
utilized  in  bridge  rectifier  service. 

32-1 1 300  Volt, 

50  Ma.  Power  Supply 

There  are  many  applications  in  the  labora- 
tory and  amateur  station  for  a simple  low 
drain  power  supply.  The  most  common  appli- 


Tl  Vi 


Figure  40 
HIGH-VOLTAGE 
POWER  SUPPLY 


COMPONENTS 

FULLLOAD 

VOLTAGE 

FULL  LOAD 
CURRENT 
(ICAS) 

Ti 

T2 

< 

1 

< 

4 

0 

1 

r 

R i 

2200- VOUT 
•POLE  TR^SFORMER 
2 KVA 

UTC 

5-71 

666-A 

20  H. 
500  MA. 
UTC  5-37 

10JJF. 
2500  V. 

75K. 
200  W, 

1900 

500  MA. 

3500-0-3500 

uTc  cc  -joe 
1 

UTC 

LS~tzi-y 

872-A 

10H. 
500  MA. 
uTccc-m 

8JLJF 
6600  V. 

200  K 
300  W, 

6000 

500  MA. 

cation  in  the  amateur  station  for  such  a sup- 
ply is  for  items  of  test  equipment  such  as  the 
type  LM  and  BC-221  frequency  meters,  for 
frequency  converters  to  be  used  in  conjunction 
with  the  station  receiver,  and  for  auxiliary 
equipment  such  as  high  selectivity  i-f  strips 
or  variable  frequency  oscillators.  Equipments 
such  as  these  may  be  operated  from  a supply 
delivering  250  to  300  volts  at  up  to  50  milli- 
amperes  of  plate  current.  A filament  source  of 
6.3  or  12  volts  may  also  be  required,  as  will 
a source  of  regulated  voltage. 

The  simple  power  supply  illustrated  in  fig- 
ures 41  and  42  is  capable  of  meeting  these 
requirements.  A two  section  capacitor  input 
filter  system  is  employed  to  provide  mini- 
mum ripple  content  and  a switch  (S2)  is 
provided  to  insert  a VR-150  regulator  tube 
in  the  circuit  to  provide  150  volts  at  ap- 
proximately 3 5 milliamperes. 

A separate  6.3  volt  filament  transformer  is 
connected  in  series  with  the  6.3  volt  wind- 
ing of  power  transformer  Ti  to  provide  12.6 
volts  at  1.2  amperes  for  operating  "12  volt" 
tubes  or  a string  of  two  "6  volt”  tubes  con- 
nected in  series.  The  secondary  winding  of 
T2  must  be  polarized  correctly  to  provide  12.6 
volts  across  the  two  windings. 

Resistor  Ri  must  be  adjusted  to  "fire”  the 
voltage  regulator  tube.  It  should  be  adjusted 
so  that  approximately  15  milliamperes  pass 
through  the  tube.  For  maximum  permissible 
current  drain  from  the  high  voltage  tap  of 
the  supply  the  VR  tube  should  be  switched 
out  of  the  circuit. 

32-12  500  Volf, 

200  Milliampere 
Power  Supply 

The  availability  of  inexpensive  "TV-type” 


Figure  41 

TYPICAL  LIGHT  DUTY 

POWER  SUPPLY 

This  300  volt,  SO  milliampere  power  supply 
may  be  used  to  run  signal  generators,  fre- 
quency meters,  small  receivers,  etc.  Switch  S2 
(see  schematic)  is  placed  on  the  rear  of  the 
chassis  near  the  line  cord. 


components  makes  possible  the  construction 
of  a sturdy  100  watt  power  supply  at  a rela- 
tively modest  price.  This  power  unit  is  suit- 
able for  operating  a 50  watt  phone/c-w  trans- 
mitter, a small  modulator,  or  a 100  watt  side- 
band linear  amplifier.  Maximum  intermittent 
output  rating  is  500  volts  at  200  milliamperes 
and  6.3  volts  at  7 amperes. 

The  circuit  of  the  supply  is  shown  in  fig- 
ure 44  and  is  designed  around  a voltage- 
doubler  type  power  transformer  {Stancor 
P-8336)  having  a 117  volt,  280  milliam- 
pere secondary  and  three  6.3  volt  filament 
windings.  The  transformer  has  a 140  watt 
core  and  is  conservatively  rated.  In  quadruplet 
service,  100  watts  of  power  may  be  taken 
from  the  high  voltage  secondary  winding  pro- 
vided the  filament  current  drain  is  held  to 
40  watts.  Under  these  limitations,  the  trans- 
former can  run  for  an  hour  or  so  before 
the  core  becomes  warm  to  the  touch. 

Four  replacement-type  selenium  rectifiers 
are  used  in  a simple  quadruplet  circuit.  Al- 
though 500  ma.  rectifiers  are  shown,  200  ma. 
units  will  work  as  well.  Five  ohm  surge  re- 
sistors are  employed  to  limit  the  starting  cur- 
rent of  the  filter  system.  Regulation  of  the 


Figure  42 

SCHEMATIC,  LIGHT  DUTY  SUPPLY 

T} — 350  - 0 - 350  volts  at  50  milliamperes,  5 
volts  at  2 amperes,  6.3  volts  at  3 amperes, 
"replacement*'  transformer. 

Ti — 6.3  volts  at  1.2  amp.  Stancor  P-6734 
CH — 4.5  henry  at  50  ma.  Stancor  C-7706 


717 


Figure  43 

TOO  WATT  "ECONOMY  MODEL" 
POWER  SUPPLY 

Employing  a TV-type  voltage  doubler  trans- 
former and  inexpensive  selenium  rectifiers, 
this  supply  delivers  500  volts  at  200  milli- 
amperes  with  good  regulation. 

The  selenium  rectifiers  are  mounted  on  a 
long  6-32  bolt  fastened  to  two  upright  posts 
mounted  at  each  front  corner  of  the  chassis. 


supply  is  good,  the  voltage  dropping  from 
600  volts  at  no  load  to  500  volts  at  maximum 
current  drain. 

The  supply  is  built  upon  a steel  chassis 
measuring  9”x6"x2”.  The  filter  capacitors 
ate  mounted  in  the  under-chassis  aiea  on 
phenolic  tie-point  strips.  All  transformer  leads 
are  left  their  full  length  so  that  the  trans- 
former will  be  in  usable  shape  in  the  event 
it  is  eventually  used  in  a different  piece  of 
equipment. 

32-13  1500  Volt, 

425  Milliompere 
Power  Supply 

One  of  the  most  popular  and  also  one  of 
the  most  convenient  power  ranges  for  ama- 
teur equipment  is  that  which  can  be  supplied 
from  a 1500  volt  power  unit  with  a current 
capability  of  about  400  ma.  The  r-f  amplifier 
of  an  A-M  phone  transmitter  ( 1 500  volts 
at  250  ma.)  capable  of  375  watts  input  and 
its  companion  modulator  (1500  volts  at  20- 
200  ma.)  can  both  be  run  from  a supply  of 
this  rating.  The  use  of  this  supply  for  SSB 
work  will  permit  a p.e.p.  of  about  600  watts 
(1500  volts  at  400  ma. ) with  the  new  low 
voltage,  high  current  RCA  7094  tetrodes. 

This  voltage  will  be  found  to  be  very  eco- 
nomical when  the  cost  of  power  supply  com- 
ponents is  computed.  A jump  in  supply  volt- 
age to  2000  will  almost  double  the  cost  of 
the  various  components.  Unless  full  kilowatt 
operation  Is  intended,  1500  volts  is  a very 
convenient  and  relatively  economical  compro- 
mise voltage. 


5 100JUF  r u 1 

10W  150V.  SR4 


^ t ITt  M'*'t 
>oV.  >iow  i 


ff-6 

&Tv, 


SRaiz^eojjF 
3S0V. 


'SRe: 


: SOUP 
350V. 


^50  K 

fiowl 


6.3  V.  AT  9 A. 


REGULATION 

LOAD(MA) 

VOLTS 

0 

600 

100 

550 

200 

500 

d-SOOV. 

AT 

200  MA. 


6.3  V. 
6.3  V. 


Figure  44 

SCHEMATIC,  ECONOMY  SUPPLY 

Tj — 117  volt  secondary  rated  at  280  ma.  for 
doubler  service.  Three  6.3  volt  filament 
windings.  Stancor  P-8336. 

CHi — 8.5  henry  at  200  ma.  Stancor  C-1721. 
SRi-SRj, — 500  mo.  se/en/um  rectifier.  Sarkes- 
Tarzian  or  equivalent. 


The  schematic  of  a typical  supply  is  shown 
in  figure  46.  Primary  power  source  may  be 
either  115  or  230  volts,  the  latter  providing 
slightly  better  power  supply  regulation.  The 
two  transformer  primaries  are  connected  in 
series  for  230  volt  operation,  or  in  parallel 
for  115  volt  operation.  In  addition  the  pri- 
maries may  be  connected  in  series  for  half- 
voltage operation  on  115  volts  as  shown.  The 
supply  provides  750  volts  at  400  ma.  under 
this  operating  condition. 

For  optimum  dynamic  voltage  regulation 
under  varying  loads  such  as  imposed  by  side- 
band or  class  B modulator  equipment  the 
output  filter  capacitor  of  the  supply  should 


Figure  45 

UNDER-CHASSIS  VIEW 
OF  ECONOMY  SUPPLY 

Two  20  /ifd.,  450  volt  capacitors  are  connected 
in  parallel  for  each  80  iifd.  unit. 


718  Power  Supplies 


THE  RADIO 


RYi  866 


Figure  46 

SCHEMATIC,  1500  VOLT,  425  MILLIAMPERE  SUPPLY 

Ti — 7710-0  - 7770  volis,  425  ma.  115/230  volt  primary.  Chicago  P-7572 
Tt — 2.5  voU^  70  ampere,  70  KV  insulation.  Stancor  P-3060 
CHt—~6  henry  at  300  ma.,  CCS.  Chicago  P-63 
RFC — ''Hash*'  suppression  choke.  Milten  77066 


be  as  large  as  is  practical.  Occasionally  60 
gfd.,  2000  volt  capacitors  can  be  picked  up 
on  the  "surplus”  market  for  a few  dollars, 
although  their  new  price  would  give  most 
amateurs  pause  for  thought.  An  inexpensive 
and  reliable  substitute  may  be  made  up  of  a 
group  of  replacement-type  tubular  electrolytic 
capacitors  connected  in  series-parallel  as  shown 
in  the  schematic.  Eight  30  gfd.,  450  volt 
cap>acitots  connected  in  series  parallel  will 
provide  an  effective  value  of  15  gfd.,  at  a 
working  voltage  of  1800.  This  is  the  mini- 
mum value  suitable  for  sideband  operation. 
Sixteen  capacitors  will  provide  30  gfd.,  at 
1800  volts. 

A power  supply  of  this  type  should  be 
built  upon  a heavy  steel  chassis,  and  all  wir- 
ing must  be  done  with  10,000  volt  TV-type 
plastic  insulated  wire.  R-F  chokes  should  be 
placed  in  the  plate  leads  of  the  mercury  vapor 
rectifiers  as  shown,  to  reduce  the  tendency 
these  tubes  have  of  breaking  into  oscillation 
over  a portion  of  the  operating  cycle.  Oscil- 
lation of  this  type  will  produce  a 120  cycle 
"buzz”  on  the  sidebands  of  the  signal.  The 
parasitic  is  eliminated  by  the  use  of  the 
chokes. 

32-14  A Dual  Volfage 
TransmiHer  Supply 

The  majority  of  high  voltage  transformers 
have  tapped  secondary  windings,  similar  to 
the  transformer  shown  in  the  schematic  of 


figure  47.  Separate  rectifier  and  filter  systems 
may  be  used  with  the  transformer  to  provide 
two  different  output  voltages  provided  the 
total  wattage  drain  from  the  supply  does  not 
exceed  the  wattage  rating  of  the  transformer. 
The  drain  may  be  divided  between  the  two 
supply  systems  in  any  manner  desired.  The 
intermittent  rating  of  T2  (figure  47)  is  750 
watts  and  the  continuous  duty  rating  is  600 
watts.  Under  CCS  rating,  the  supply  can  pro- 
vide (for  example)  2000  volts  at  I6O  ma. 
for  the  operation  of  an  813  r-f  amplifier  at 
320  watts  input,  and  1750  volts  at  20-150 
ma,  for  the  operation  of  811-A  class  B modu- 
lators. Under  intermittent  duty  rating,  the 
813  amplifier  can  run  at  400  watts  input 
(phone)  and  500  watts  (c-w)  without  over- 
loading the  supply. 

A remote  switch  is  used  to  energize  the 
plate  circuit  relay  of  the  supply.  An  auxiliary 
antenna  relay  is  also  operated  by  the  "trans- 
mit” switch. 

32-15  A Kilowatt' 

Power  Supply 

Shown  in  figure  48  is  the  schematic  of  a 
power  supply  capable  of  delivering  2500  volts 
at  a continuous  current  drain  of  500  milli- 
amperes,  or  700  milliamperes  with  an  inter- 
mittent load.  The  supply  is  designed  to  power 
a kilowatt  amplifier  operating  at  2500  volts 
and  400  ma.,  in  conjunction  with  a 500  watt 
modulator  operating  at  2500  volts  at  a vary- 


HANDBOOK 


Special  Supplies  719 


Figure  47 

DUAL  VOLTAGE  POWER  SUPPLY 
T 2,  Ts — 2.5  volts  at  5 amperes.  Stancor  P-6733 

Ti — 2400-2700  vo71s  each  side  center  tap  at  375  mo.  /CAS.  Stancor  P-8032 
CHi — 3 to  77  henry,  300  mo.  Stancor  C-7403 
CHt — 8 henry,  300  mo.,  Stancor  C-7473 
R»— 70K.,  700  wott 

RFCi,  RFCi — "Hash"  suppression  choke.  J.  W.  Miller  7865  twin  chokes  (2  rea.) 
RYi— SPST  re/oy,  IIS  volt  coil. 


ing  current  drain  of  50-300  ma.  Specifically, 
the  supply  is  employed  with  a transmitter 
having  a pair  of  4-2 50 A tetrode  tubes  in  the 
class  C stage,  and  a pair  of  810  modulator 
tubes.  For  sideband  work,  the  supply  may  be 
used  to  power  a 1750  watt  p.e.p.  linear  ampli- 
fier, such  as  the  4CX-1000A  amplifier  shown 
in  an  earlier  chapter. 

Because  the  total  weight  of  the  components 
is  over  150  pounds,  the  supply  should  be 
built  directly  on  the  bottom  of  a relay  rack 
instead  of  upon  a steel  chassis. 

The  t-f  hash  suppression  chokes  RFG  and 
RFC:  are  fastened  directly  to  the  high  voltage 
terminals  of  the  plate  transformer.  The  two 
872- A rectifier  tubes  ate  so  located  that  the 
leads  from  the  r-f  chokes  to  the  plate  caps 
are  only  about  three  inches  long. 

A 0.15  gfd.,  5000  volt  paper  capacitor  is 
used  to  resonate  the  filter  choke  to  approxi- 
mately 120  cycles  at  a bleeder  current  of  25 
milliamperes.  When  full  load  current  is 
drawn,  the  inductance  of  the  filter  choke 
drops,  detuning  the  parallel  resonant  circuit. 
Improved  voltage  regulation  is  gained  by  this 
action;  the  no  load  voltage  increases  only 
200  volts  over  the  full  load  voltage. 


872A'5 


Tl -2900-0-2900  VOLTS  AT  700  MA.,  ICAS,  CHICAGO  P-2126 
T2-5  VOLTS,  10  AMP.,  CHICAGO  F-S10H 
CHi-6  HENRIES,  700  MA.  CHICAGO  R-67 
Cl-0.15JUF,  5000-VOLT. 

C2- THREE  4-UF  3000-VOLT 
Rl- 100,000  OHMS.  200-Vt/ATT. 

R2-ELEVEN  0.5  MEG.  2-WATT  RESISTORS  IN  SERIES 

RYl  -DPST  RELAY,  110  V.  COIL.  20  A.  CONTACTS.  POTTER  A BRUMFIELD 

PR7A 

RFCl,  RFC2-  HASH  FILTER.  J.W.  MILLER  CO-  H”  7666 


Figure  48 

HIGH  VOLTAGE  POWER  SUPPLY 


CHAPTER  THIRTY-THREE 


With  a few  possible  exceptions,  such  as 
fixed  air  capacitors,  neutralizing  capacitors 
and  transmitting  coils,  it  hardly  pays  one  to 
attempt  to  build  the  components  required  for 
the  construction  of  an  amateur  transmitter. 
This  is  especially  true  when  the  parts  are  of 
the  type  used  in  construction  and  replacement 
work  on  broadcast  receivers,  as  mass  produc- 
tion has  made  these  parts  very  inexpensive. 

Transmitters  Those  who  have  and  wish  to 
spend  the  necessary  time  can 
effect  considerable  monetary  saving  in  their 
transmitters  by  building  them  from  the  com- 
ponent parts.  The  necessary  data  is  given 
in  the  construction  chapters  of  this  handbook. 

To  many  builders,  the  construction  is  as 
fascinating  as  the  operation  of  the  finished 
transmitter;  in  fact,  many  amateurs  get  so 
much  satisfaction  out  of  building  a well-per- 
forming piece  of  equipment  that  they  spend 
more  time  constructing  and  rebuilding  equip- 
ment than  they  do  operating  the  equipment 
on  the  air. 


33-1  Tools 

Beautiful  work  can  be  done  with  metal 
chassis  and  panels  with  the  help  of  only  a 
few  inexpensive  tools.  The  time  required  for 
construction,  however,  will  be  greatly  reduced 
if  a fairly  complete  assortment  of  metal-work- 
ing tools  is  available.  Thus,  while  an  array  of 
tools  will  speed  up  the  work,  excellent  results 
may  be  accomplished  with  few  tools,  if  one 
has  the  time  and  patience. 

The  investment  one  is  justified  in  making 
in  tools  is  dependent  upon  several  factors.  If 
you  like  to  tinker,  there  are  many  tools  use- 
ful in  radio  construction  that  you  would  prob- 
ably buy  anyway,  or  perhaps  already  have, 
such  as  screwdrivers,  hammer,  saws,  square, 
vise,  files,  etc.  This  means  that  the  money 
taken  for  tools  from  your  radio  budget  can  be 
used  to  buy  the  more  specialized  tools,  such 
as  socket  punches  or  hole  saws,  taps  and  dies, 
etc. 

The  amount  of  construction  work  one  does 
determines  whether  buying  a large  assortment 


720 


Tools  721 -A 


of  tools  is  an  economical  move.  It  also  deter- 
mines if  one  should  buy  the  less  expensive 
type  offered  at  surprisingly  lov/  prices  by  the 
familiar  mail  order  houses,  "five  and  ten” 
stores  and  chain  auto-supply  stores,  or  whether 
one  should  spend  more  money  and  get  first- 
grade  tools.  The  latter  cost  considerably  more 
and  work  but  little  better  when  new,  but  will 
outlast  several  sets  of  the  cheaper  tools. 
Therefore  they  are  a wise  investment  for  the 
experimenter  who  does  lots  of  construction 
work  (if  he  can  afford  the  initial  cash  outlay). 
The  amateur  who  constructs  only  an  occasional 
piece  of  apparatus  need  not  be  so  concerned 
with  tool  life,  as  even  the  cheaper  grade  tools 
will  last  him  several  years,  if  they  are  given 
proper  care. 

The  hand  tools  and  materials  in  the  accom- 
panying lists  will  be  found  very  useful  around 
the  home  workshop.  Materials  not  listed  but 
ordinarily  used,  such  as  paint,  can  best  be 
purchased  as  required  for  each  individual  job. 

ESSENTIAL  HAND  TOOLS  AND 
MATERIALS 

1 Good  electric  soldering  iron,  about  100 
watts, 

1 Spool  rosin-core  wire  solder 
1 Each  large,  medium,  small,  and  midget 
screwdrivers 

1 Good  hand  drill  (eggbeater  type),  prefer- 
ably two-speed 

1 Pair  tegular  pliers,  6 inch 
1 Pair  long  nose  pliers,  6 inch 
1 Pair  cutting  pliers  (diagonals),  5 inch  or 
6 inch 

1 IVs-inch  tube-socket  punch 
1 "Boy  Scout”  knife 

1 Combination  square  and  steel  rule,  1 foot 
1 Yardstick  or  steel  pushrule 
1 Scratch  awl  or  ice  pick  scribe 
1 Center  punch 

1 Do2en  or  more  assorted  round  shank  drills 
(as  many  as  you  can  afford  between  no. 
50  and  14  or  % inch,  depending  upon 
size  of  hand  drill  chuck) 

1 Combination  oil  stone 
Light  machine  oil  (in  squirt  can) 

Friction  tape 
1 Hacksaw  and  blades 
1 Medium  file  and  handle 

1 Cold  chisel  (V2  inch  tip) 

1 Wrench  for  socket  punch 
1 Hammer 


HIGHLY  DESIRABLE  HAND  TOOLS 
AND  MATERIALS 
1 Bench  vise  (jaws  at  least  3 inch) 

1 Spool  plain  wire  solder 
I Carpenter’s  brace,  ratchet  type 
1 Square-shank  countersink  bit 
1 Square-shank  taper  reamer,  small 
1 Square-shank  taper  reamer,  large  (the  two 
reamers  should  overlap;  I/2  inch  and  % 
inch  size  will  usually  be  suitable) 

1 % inch  tube-socket  punch  ( for  electrolytic 

capacitors ) 

1 1-3/16  inch  tube-socket  punch 

1 5^ -inch  tube-socket  punch 

1 Adjustable  circle  cutter  for  holes  to  3 inch 
1 Set  small,  inexpensive,  open-end  wrenches 
1 Pair  tin  shears,  10  or  12  inch 
1 Wood  chisel  (V2  inch  tip) 

1 Pair  wing  dividers 
1 Coarse  mill  file,  flat  12  inch 
1 Coarse  bastard  file,  round,  I/2  or  inch 
1 Set  alien  and  spline-head  wrenches 
6 or  8 Assorted  small  files;  round,  half-round 
or  triangular,  flat,  square,  rat-tail 
4 Small  "C”  clamps 
Steel  wool,  coarse  and  fine 
Sandpaper  and  emery  cloth,  coarse,  medium, 
and  fine 
Duco  cement 
File  brush 

USEFUL  BUT  NOT  ESSENTIAL  TOOLS 
AND  MATERIALS 

1 Jig  or  scroll  saw  (small)  with  metal-cut- 
ting blades 

1 Small  wood  saw  (crosscut  teeth) 

1 Each  square-shank  drills:  7/16,  and  V2 

inch 

1 Tap  and  die  outfit  for  6-32,  8-32,  10-32 
and  10-24  machine  screw  threads 
4 Medium  size  "C”  clamps 
Lard  oil  (in  squirt  can) 

Kerosene 
Empire  cloth 

Clear  lacquer  ("industrial”  grade) 

Lacquer  thinner 
Dusting  brush 
Paint  brushes 

Sheet  celluloid,  Lucite,  or  polystyrene 
1 Carpenter’s  plane 

1 Each  "Spintite”  wrenches,  |4,  5/16,  11/32 
to  fit  the  standard  6-32  and  8-32  nuts 
used  in  radio  work 

1 Screwdriver  for  recessed  head  type  screws 

The  foregoing  assortment  assumes  that  the 
constructor  does  not  want  to  invest  in  the 
more  expensive  power  tools,  such  as  drill  press. 


THE  RADIO 


Figure  1 

SOFT  ALUMINUM 
SHEET  MAY  BE  CUT 
WITH  HEAVY 
KITCHEN  SHEARS 


grinding  head,  etc.  If  power  equipment  is  pur- 
chased, obviously  some  of  the  hand  tools  and 
accessories  listed  will  be  superfluous.  A drill 
press  greatly  facilitates  construction  work, 
arid  it  is  unfortunate  that  a good  one  costs  as 
much  as  a small  transmitter. 

Not  listed  in  the  table  are  several  special- 
purpose  radio  tools  which  are  somewhat  of  a 
luxury,  but  are  nevertheless  quite  handy,  such 
as  various  around-the-corner  screwdrivers  and 
wrenches,  special  soldering  iron  tips,  etc. 
These  can  be  found  in  the  larger  radio  parts 
stores  and  are  usually  listed  in  their  mail  or- 
der catalogs. 

If  it  is  contemplated  to  use  the  newer  and 
very  popular  mlniamre  series  of  tubes  (6AK5, 
6C4,  6BA6,  etc.)  in  the  construction  of  equip- 
ment certain  additional  tools  will  be  required 
to  mount  the  smaller  components.  Miniature 


tube  sockets  mount  in  a %-inch  hole,  while 
9-pin  sockets  mount  in  a %-inch  hole.  Green- 
lee socket  punches  can  be  obtained  in  these 
sixes,  or  a smaller  hole  may  be  reamed  to  the 
proper  size.  Needless  to  say,  the  punch  is 
much  the  more  satisfactory  solution.  Mounting 
screws  for  miniature  sockets  are  usually  of 
the  4-40  size. 

Metal  Chossis  Though  quite  a few  more  tools 
and  considerably  more  time 
will  be  required  for  metal  chassis  construction, 
much  neater  and  more  satisfaaory  equipment 
can  be  built  by  mounting  the  parts  on  sheet 
metal  chassis  instead  of  breadboards.  This  type 
of  construction  is  necessary  when  shielding  of 
the  appartus  is  required.  A front  panel  and  a 
back  shield  minimize  the  danger  of  shock  and 
complete  the  shielding  of  the  enclosure. 


Figure  2 

CONVENTIONAL 
WOOD  EXPANSION 
BIT  IS  EFFECTIVE  IN 
DRILLING  SOCKET 
HOLES  IN  REYNOLDS 
DO-IT-YOURSELF 
ALUMINUM 


HANDBOOK 


Material  723-A 


Figure  3 

SOFT  ALUMINUM 
TUBING  MAY  BE 
BENT  AROUND 
WOODEN  FORM 
BLOCKS.  TO  PREVENT 
THE  TUBE  FROM 
COLLAPSING  ON 
SHARP  BENDS,  IT  IS 
PACKED  WITH 
WET  SAND. 


33-2  The  Material 

Elearonk  equipment  may  be  built  upon 
foundation  of  wood,  steel  or  aluminum.  The 
choice  of  foundation  material  is  governed  by  the 
requirements  of  the  electrical  circuit,  the  weight 
of  the  components  of  the  assembly,  and  the  fin- 
ancial cost  of  the  project  when  balanced  against 
the  pocketbook  contents  of  the  constructor. 

Breodboard  The  simplest  method  of  con- 
structing equipment  is  to  lay  it 
out  in  breadboard  fashion,  which  consists  of 
fastening  the  various  components  to  a board 
of  suitable  size  with  wood  screws  or  machine 
bolts,  arranging  the  parts  so  that  important 
leads  will  be  as  short  as  possible. 

Breadboard  construction  is  suitable  for  test- 
ing an  experimental  layout,  or  sometimes  for 
assembling  an  experimental  unit  of  test  equip- 
ment. But  no  permanent  item  of  station  equip- 
ment should  be  left  in  the  breadboard  form. 
Breadboard  construction  is  dangerous,  since 
components  carrying  dangerous  voltages  are 
left  exposed.  Also,  breadboard  construction  is 
never  suitable  for  any  r-f  portion  of  a trans- 
mitter, since  it  would  be  substantially  impos- 
sible to  shield  such  an  item  of  equipment  for 
the  elimination  of  TVI  resulting  from  har- 
monic radiation. 


Figure  4 

A WOODWORKING 
PLANE  MAY  BE  USED 
TO  SMOOTH  OR 
TRIM  THE  EDGES  OF 
REYNOLDS 
DO-IT-YOURSELF 
ALUMINUM  STOCK. 


Dish  type  construction  is  practically  the 
same  as  metal  chassis  construction,  the  main 
difference  lying  in  the  manner  in  which  the 
chassis  is  fastened  to  the  panel. 

Special  For  high-powered  t-f  stages, 

Fromeworkt  many  amateur  constructors  pre- 
fer to  discard  the  more  conven- 
tional types  of  construction  and  employ  in- 
stead special  metal  frameworks  and  brackets 
which  they  design  specially  for  the  parts  which 
they  intend  to  use.  These  ate  usually  arranged 
to  give  the  shortest  possible  r-f  leads  and  to 
fasten  directly  behind  a relay  rack  panel  by 
means  of  a few  bolts,  with  the  control  shafts 


724-A  Workshop  Practice 


THE  RADIO 


Figure  5 

INEXPENSIVE  OPERATING  DESK  MADE 
FROM  ALUMINUM  ANGLE  STOCK,  PLY- 
WOOD AND  A FLUSH-TYPE  DOOR 


projecting  through  corresponding  holes  in  the 
panel. 

Working  with  The  necessity  of  employing 
Aluminum  "electrically  tight  enclosures” 
for  the  containment  of  TVI- 
producing  harmonics  has  led  to  the  general 
use  of  aluminum  for  chassis,  panel,  and  en- 
closure construction.  If  the  proper  type  of 
aluminum  material  is  used,  it  may  be  cut  and 
worked  with  the  usual  woodworking  tools 
found  in  the  home  shop.  Hard,  brittle  alumi- 
num alloys  such  as  24ST  and  6 1ST  should  be 
avoided,  and  the  softer  materials  such  as  2S 
or  1/2  H should  be  employed. 

A new  market  product  is  Reynold’s  Do-it- 
yourself  aluminum,  which  is  being  distributed 
on  a nationwide  basis  through  hardware  stores, 
lumber  yards  and  building  material  outlets. 
This  material  is  an  alloy  which  is  temper  se- 
lected for  easy  working  with  ordinary  tools. 
Aluminum  sheet,  bar  and  angle  stock  may  be 
obtained,  as  well  as  perforated  sheets  for  ven- 
tilated enclosures. 

Figures  1 through  4 illustrate  how  this  soft 
material  may  be  cut  and  worked  with  ordinary 
shop  tools,  and  fig.  5 shows  a simple  operating 
desk  that  may  be  made  from  aluminum  angle 
stock,  plywood  and  a flush-type  six  foot  door. 


Figure  6 

TVI  ENCLOSURE  MADE  FROM 
SINGLE  SHEET  OF 
PERFORATED  ALUMINUM 

Reynolds  Metal  Co.  *"Do-ii’‘yourself**  aluminum 
sheet  may  be  cut  and  folded  to  form  TVI- 
proof  enclosure.  One~half  inch  lip  on  edges  Is 
bolted  to  center  section  with  6*32  machine 
screws. 


33-3  TVI-Proof 
Enclosures 

Armed  with  a right-angle  square,  a tin-snips 
and  a straight  edge,  the  home  constructor  will 
find  the  assembly  of  aluminum  enclosures  an 
easy  task.  This  section  will  show  simple  con- 
struction methods,  and  short  cuts  in  producing 
enclosures. 

The  simplest  type  of  aluminum  enclosure 
is  that  formed  from  a single  sheet  of  per- 
forated material  as  shown  in  figure  6.  The 
top,  sides,  and  back  of  the  enclosure  are  of 
one  piece,  complete  with  folds  that  permit 
the  formed  enclosure  to  be  bolted  together 
along  the  edges.  The  top  area  of  the  enclosure 
should  match  the  area  of  the  chassis  to  en- 
sure a close  fit.  The  front  edge  of  the  en- 
closure is  attached  to  aluminum  angle  strips 
that  are  bolted  to  the  front  panel  of  the 
unit;  the  sides  and  back  can  either  be  bolted 
to  matching  angle  strips  affixed  to  the  chassis, 
or  may  simply  be  attached  to  the  edge  of  the 
chassis  with  self-tapping  sheet  metal  screws. 
Enclosures  of  this  type  are  used  on  the  all- 
band transmitter  described  in  chapter  31. 

A more  sophisticated  enclosure  is  shown 
in  figure  7.  In  this  assembly  aluminum  angle 
stock  is  cut  to  length  to  form  a framework 
upon  which  the  individual  sides,  back,  and 
top  of  the  enclosure  are  bolted.  For  greatest 
strength,  small  aluminum  gusset  plates  should 
be  affixed  in  each  corner  of  the  enclosure. 
The  complete  assembly  may  be  held  together 
by  no.  6 sheet  metal  screws. 


HANDBOOK 


Openings  725-A 


Regardless  of  the  type  of  enclosure  to  be 
made,  care  should  be  taken  to  ensure  that  all 
joints  are  square.  Do  not  assume  that  all  pre- 
fabricated chassis  and  panel  are  absolutely  true 
and  square.  Check  them  before  you  start  to 
form  your  shield  as  any  dimensional  errors  in 
the  foundation  will  cause  endless  patching  and 
cutting  after  your  enclosure  is  bolted  together. 
Finally,  be  sure  that  paint  is  remcwed  from 
the  panel  and  chassis  at  the  point  the  en- 
closure attaches  to  the  foundation.  A clean, 
metallic  contact  along  the  seam  is  required 
for  maximum  harmonic  suppression. 

33-4  Enclosure 
Openings 

openings  into  shielded  enclosures  may  be 
made  simply  by  covering  them  by  a piece  of 
shielding  held  in  place  by  sheet  metal  screws. 

Openings  through  vertical  panels,  however, 
usually  require  a bit  more  attention  to  pre- 
vent leakage  of  harmonic  energy  through  the 
crack  of  the  door  which  is  supposed  to  seal 
the  opening.  A simple  way  to  provide  a panel 
opening  is  to  employ  the  Bud  ventilated  door 
rack  panel  model  PS-814  or  813.  The  grille 
opening  in  this  panel  has  holes  small  enough 
in  area  to  prevent  serious  harmonic  leakage. 
The  actual  door  opening,  however,  does  not 
seal  tightly  enough  to  be  called  TVI-proof. 
In  areas  of  high  TV  signal  strength  where 
a minimum  of  operation  on  28  Me.  is  con- 
templated, the  door  is  satisfactory  as  is.  To 
accomplish  more  complete  harmonic  suppres- 
sion, the  edges  of  the  opening  should  be  lined 
with  preformed  contact  finger  stock  manufac- 
tured by  Eitel-McCullough,  Inc.,  of  San  Bruno, 
Calif.  Eimac  finger  stock  is  an  excellent  means 
of  providing  good  contact  continuity  when 
using  components  with  adjustable  or  moving 
contact  surfaces,  or  in  acting  as  electrical 
"weatherstrip”  around  access  doors  in  enclo- 
sures. Harmonic  leakage  through  such  a sealed 
opening  is  reduced  to  a negligible  level.  The 
mating  surface  to  the  finger  stock  should  be 
free  of  paint,  and  should  provide  a good  elec- 
trical connection  to  the  stock. 

A second  method  of  re-establishing  elec- 
trical continuity  across  an  access  port  is  to 
employ  Metex  shielding  around  the  mating 
edges  of  the  opening.  Metex  is  a flexible  knit- 
ted wire  mesh  which  may  be  obtained  in 
various  sizes  and  shapes.  This  r-f  gasket  ma- 
terial is  produced  by  Metal  Textile  Corp., 
Roselle,  N.J.  Metex  is  both  flexible  and  resili- 
ent and  conforms  to  irregularities  in  mating 
surfaces  with  a minimum  of  closing  pressure. 


Figure  7 

TVI-PROOF  ENCLOSURE  BUILT  OF 
PERFORATED  ALUMINUM  SHEET 
AND  ANGLE  STOCK 


33-5  Summation 

of  the  Problem 

The  creation  of  r-f  energy  is  accompanied 
by  harmonic  generation  and  leakage  of  funda- 
mental and  harmonic  energy  from  the  generator 
source.  For  practical  purposes,  radio  frequen- 
cy power  may  be  considered  as  a form  of  both 
electrical  and  r-f  energy.  As  electrical  energy, 
it  will  travel  along  any  convenient  conductor. 
As  r-f  energy,  it  will  radiate  directly  from  the 
source  or  from  any  conductor  connected  to  the 
source.  In  view  of  this  "dual  personality”  of 
r-f  emanations,  there  is  no  panacea  for  all 
forms  of  r-f  energy  leakage.  The  cure  involves 
both  filtering  and  shielding;  one  to  block  the 
paths  of  conducted  energy,  the  other  to  pre- 
vent the  leakage  of  radiated  energy.  The  proper 
combination  of  filtering  and  shielding  can  re- 
duce the  radiation  of  harmonic  energy  from  a 
signal  source  some  80  decibels.  In  most  cases, 
this  is  sufficient  to  eliminate  interference  caused 
by  the  generation  of  undesirable  harmonics. 


726-A  Workshop  Practice 


THE  RADIO 


33-6  Construction 
Practice 

Chassis  The  chassis  first  should  be  covered 
Layout  with  a layer  of  wrapping  paper, 
which  is  drawn  tightly  down  on 
all  sides  and  fastened  with  scotch  tape.  This 
allows  any  number  of  measurement  lines  and 
hole  centers  to  be  spotted  in  the  correct  po- 
sitions without  making  any  marks  on  the 
chassis  itself.  Place  on  it  the  parts  to  be 
mounted  and  play  a game  of  chess  with  them, 
trying  different  arrangements  until  all  the  grid 
and  plate  leads  are  made  as  short  as  possible, 
tubes  are  clear  of  coil  fields,  r-f  chokes  are  in 
safe  positions,  etc.  Remember,  especially  if 
you  are  going  to  use  a panel,  that  a good  me- 
chanical layout  often  can  accompany  sound 
electrical  design,  but  that  the  electrical  de- 
sign should  be  given  first  consideration. 

All  too  often  parts  are  grouped  to  give  a 
symmetrical  panel,  irrespective  of  the  arrange- 
ment behind.  When  a satisfactory  arrangement 
has  been  reached,  the  mounting  holes  may  be 
marked.  The  same  procedure  now  must  be 
followed  for  the  underside,  always  being  care- 
ful to  see  that  there  are  no  clashes  between 
the  two  (that  no  top  mounting  screws  come 
down  into  the  middle  of  a paper  capacitor  on 
the  underside,  that  the  variable  capacitor  rotors 
do  not  hit  anything  when  turned,  etc. ) . 

When  all  the  holes  have  been  spotted,  they 
should  be  center-punched  through  the  paper 
into  the  chassis.  Don’t  forget  to  spot  holes  for 
leads  which  must  also  come  through  the 
chassis. 

For  transformers  which  have  lugs  on  the 
bottoms,  tbe  clearance  holes  may  be  spotted 
by  pressing  the  transformer  on  a piece  of  pa- 
per to  obtain  impressions,  which  may  then 
be  transferred  to  the  chassis. 

Punching  In  cutting  socket  holes,  one  can 
use  either  a fly-cutter  or  socket 
punches.  These  punches  are  easy  to  operate 
and  only  a few  precautions  are  necessary.  The 
guide  pin  should  fit  snugly  in  the  guide  hole. 
This  increases  the  accuracy  of  location  of  the 
socket.  If  this  is  not  of  great  importance,  one 
may  well  use  a drill  of  1/32  inch  larger  di- 
ameter than  the  guide  pin.  Some  of  the  punches 
will  operate  without  guide  holes,  but  the  latter 
always  make  the  punching  operations  simpler 
and  easier.  The  only  other  precaution  is  to  be 
sure  the  work  is  properly  lined  up  before  ap- 
plying the  hammer.  If  this  is  not  done,  the 
punch  may  slide  sideways  when  you  strike  and 
thus  not  only  shear  the  chassis  but  also  take 


EOeOODOC 

§ 

1 ' 

1 i 

[ 1 

[ 1 

(D  ® @ 

DRILL  HOLES  SLIGHTLY  BREAKOUT  FILE 

INSIDE  DASHED  OUTLINE  PIECE  INSIDE  SMOOTH 

OF  DESIRED  HOLE.  DRILLHOLES. 

MAKING  RECTANGULAR  CUTOUT 

Figure  8 


off  part  of  the  die.  This  is  easily  avoided  by 
always  making  sure  that  the  piece  is  parallel 
to  the  faces  of  the  punch,  the  die,  and  the 
base.  The  latter  should  be  an  anvil  or  other 
solid  base  of  heavy  material. 

A punch  by  Greenlee  forces  socket  holes 
through  the  chassis  by  means  of  a screw  turned 
with  a wrench.  It  is  noiseless,  and  works  much 
more  easily  and  accurately  than  most  others. 

The  male  part  of  the  punch  should  be  placed 
in  the  vise,  cutting  edge  up  and  the  female 
portion  forced  against  the  metal  with  a wrench. 
These  punches  can  be  obtained  in  sizes  to 
accommodate  all  tube  sockets  and  even  large 
enough  to  be  used  for  meter  holes.  In  the 
octal  socket  sizes  they  require  the  use  of  a 
inch  center  hole  to  accommodate  the  bolt. 

Transformer  Cutouts  for  transformers  and 
Cutouts  chokes  are  not  so  simply  han- 
dled. After  marking  off  the  part 
to  be  cut,  drill  about  a 14 -inch  hole  on  each 
of  the  inside  corners  and  tangential  to  the 
edges.  After  burring  the  holes,  clamp  the  piece 
and  a block  of  cast  iron  or  steel  in  the  vise. 
Then,  take  your  burring  chisel  and  insert  it  in 
one  of  the  corner  holes.  Cut  out  the  metal  by 
hitting  the  chisel  with  a hammer.  The  blows 
should  be  light  and  numerous.  The  chisel  acts 
against  the  block  in  the  same  way  that  the  two 
blades  of  a pair  of  scissors  work  against  each 
other.  This  same  process  is  repeated  for  the 
other  sides.  A file  is  used  to  trim  up  the  com- 
pleted cutout. 

Another  method  is  to  drill  the  four  corner 
holes  large  enough  to  take  a hack  saw  blade, 
then  saw  instead  of  chisel.  The  four  holes  per- 
mit nice  looking  corners. 

Still  another  method  is  shown  in  figure  8. 
When  heavy  panel  steel  is  used  and  a drill 
press  or  electric  drill  is  available,  this  is  the 
most  satisfactory  method. 

Removing  In  both  drilling  and  punching,  a 
Burrs  burr  is  usually  left  on  the  work. 

There  are  three  simple  ways  of 


HANDBOOK 


Construction  Practice  727-A 


removing  these.  Perhaps  the  best  is  to  take  a 
chisel  (be  sure  it  is  one  for  use  on  metal)  and 
set  it  so  that  its  bottom  face  is  parallel  to  the 
piece.  Then  gently  tap  it  with  a hammer.  This 
usually  will  make  a clean  job  with  a little 
practice.  If  one  has  access  to  a counterbore, 
this  will  also  do  a nice  job.  A countersink  will 
work,  although  it  bevels  the  edges.  A drill  of 
several  sizes  larger  is  a much  used  arrange- 
ment. The  third  method  is  by  filing  off  the 
burr,  which  does  a good  job  but  scratches  the 
adjacent  metal  surfaces  badly. 

Mounting  There  are  two  methods  in  gen- 
Components  eral  use  for  the  fastening  of 
transformers,  chokes,  and  similar 


NUMBERED 

DRILL  SIZES 

DRILL 

NUMBER 

1 

Di- 
ameter Clears 

(in.)  Screw 

Correct  for 
Tapping 
Steel  or 
Brasst 

2 

.221 

12-24 

3 .. 

.213 

14-24 

4 . 

12-20 

5 

6 ..  . 

.204 

7 .. 

8 . 

9 . 

10«. ... 

.193 

10-32 

10-24 

11  .... 

.191 

12*.... 

.189 

13 

14  ..  . 

15  . 

14  ..  . 

.177 

17 

.173 

18* 

8-32 

19 

12-20 

20 

21*... 

.159 

22  ... 

.157 

23 

24 

25* 

.149 

26 

27  ... 

28*.... 

.140 

6-32 

29*  .. 

.136 

31 

32 

33* 

.113 

4-36  4-40 

34 

35*.... 

.110 

36 

37 

38  

.102 

39*.... 

.100 

3-48 

40  .... 

.098 

41 

42* 

.093 

43  

.089 

2-56 

44  

.086 

45* 

.082 

tUse 

bokelite 

(plastics. 

next 

and 

etc.) 

size  larger  drill  for  tapping 
similor  composition  materials 

*Sizes 

struetion 

most 

commonly  used  in 

radio  con- 

Figure  9 


pieces  of  apparatus  to  chassis  or  breadboards. 
The  first,  using  nuts  and  machine  screws,  is 
slow,  and  the  commercial  manufacturing  prac- 
tice of  using  self-tapping  screws  is  gaining 
favor.  For  the  mounting  of  small  parts  such 
as  resistors  and  capacitors,  "tie  points’’  are 
very  useful  to  gain  rigidity.  They  also  con- 
tribute materially  to  the  appearance  of  finished 
apparatus. 

Rubber  grommets  of  the  proper  size,  placed 
in  all  chassis  holes  through  which  wires  are 
to  be  passed,  will  give  a neater  appearing  job 
and  also  will  reduce  the  possibility  of  short 
circuits. 

Soldering  Making  a strong,  low-resistance 
solder  joint  does  not  mean  just 
dropping  a blob  of  solder  on  the  two  parts  to 
be  joined  and  then  hoping  that  they’ll  stick. 
There  are  several  definite  rules  that  must  be 
observed. 

All  parts  to  be  soldered  must  be  absolutely 
clears.  To  clean  a wire,  lug,  or  whatever  it  may 
be,  take  your  pocket  knife  and  scrape  it  thor- 
oughly, until  fresh  metal  is  laid  bare.  It  is  not 
enough  to  make  a few  streaks;  scrape  until  the 
part  to  be  soldered  is  bright. 

Make  a good  mechanical  joint  before  apply- 
ing any  solder.  Solder  is  intended  primarily 
to  make  a good  electrical  connection;  mechani- 
cal rigidity  should  be  obtained  by  bending  the 
wire  into  a small  hook  at  the  end  and  nipping 
it  firmly  around  the  other  part,  so  that  it  will 
hold  well  even  before  the  solder  is  applied. 

Keep  your  iron  properly  tinned.  It  is  im- 
possible to  get  the  work  hot  enough  to  take 
the  solder  properly  if  the  iron  is  dirty.  To  tin 
your  iron,  file  it,  while  hot,  on  one  side  until 
a full  surface  of  clean  metal  is  exposed.  Im- 
mediately apply  rosin  core  solder  until  a thin 
layer  flows  completely  over  the  exposed  sur- 
face. Repeat  for  the  other  faces.  Then  take  a 
clean  rag  and  wipe  off  all  excess  solder  and 
rosin.  The  iron  should  also  be  wiped  frequently 
while  the  actual  construction  is  going  on;  it 
helps  prevent  pitting  the  tip. 

Apply  the  solder  to  the  work,  not  to  the 
iron.  The  iron  should  be  held  against  the  parts 
to  be  joined  until  they  are  thoroughly  heated. 
The  solder  should  then  be  applied  against  the 
parts,  and  the  iron  should  be  held  in  place 
until  the  solder  flows  smoothly  and  envelopes 
the  work.  If  it  acts  like  water  on  a greasy 
plate,  and  forms  a ball,  the  work  is  not  suf- 
ficiently clean. 

The  completed  joint  must  be  held  perfectly 
still  until  the  solder  has  had  time  to  solidify. 
If  the  work  is  moved  before  the  solder  has  be- 


728-A  Workshop  Practice 


WINDING  COIL  ON  INSULATING  FORM 


Figure  10 


Figure  1 1 


come  completely  solid,  a "cold”  joint  will  re- 
sult. This  can  be  identified  immediately,  be- 
cause the  solder  will  have  a dull  "white”  ap- 
pearance rather  than  one  of  shiny  "silver.” 
Such  joints  tend  to  be  of  high  resistance  and 
will  very  likely  have  a bad  effect  upon  a cir- 
cuit. The  cure  is  simple,  merely  reheat  the 
joint  and  do  the  job  correctly. 

Wipe  away  all  surplus  flux  when  the  joint 
has  cooled  if  you  are  using  a paste  type  flux. 
Be  sure  it  is  non-corrosive,  and  use  it  with 
plain  (not  rosin  core)  solder. 

Finishes  If  the  apparatus  is  constructed  on 
a painted  chassis  (commonly  avail- 
able in  black  wrinkle  and  gray  wrinkle),  there 
is  no  need  for  application  of  a protective  coat- 
ing when  the  equipment  is  finished,  assuming 
that  you  are  careful  not  to  scratch  or  mar  the 
finish  while  drilling  holes  and  mounting  parts. 
However,  many  amateurs  prefer  to  use  un- 
painted (zinc  or  cadmium  plated)  chassis,  be- 
cause it  is  much  simpler  to  make  a chassis 
ground  connection  with  this  type  of  chassis. 
A thin  coat  of  clear  "linoleum”  lacquer  may 
be  applied  to  the  whole  chassis  after  the  wir- 
ing is  completed  to  retard  rusting.  In  localities 
near  the  sea  coast  it  is  a good  idea  to  lacquer 
the  various  chassis  cutouts  even  on  a painted 
chassis,  as  rust  will  get  a good  start  at  these 
points  unless  the  metal  is  protected  where  the 
drill  or  saw  has  exposed  it.  If  too  thick  a coat 
is  applied,  the  lacquer  will  tend  to  peel.  It 
may  be  thinned  with  lacquer  thinner  to  permit 
application  of  a light  coat.  A thin  coat  will 
adhere  to  any  clean  metal  surface  that  is  not 
too  shiny. 

An  attractive  dull  gloss  finish,  almost  vel- 
vety can  be  put  on  aluminum  by  sand-blasting 
it  with  a very  weak  blast  and  fine  particles 
and  then  lacquering  it.  Soaking  the  aluminum 
in  a solution  of  lye  produces  somewhat  the 
same  effect  as  a fine  grain  sand  blast. 


There  are  also  several  brands  of  dull  gloss 
black  enamels  on  the  market  which  adhere 
well  to  metals  and  make  a nice  appearance. 
Airdrying  wrinkle  finishes  are  sometimes  suc- 
cessful, but  a bake  job  is  usually  far  better. 
Wrinkle  finishes,  properly  applied,  are  very 
durable  and  are  pleasing  to  the  eye.  If  you 
live  in  a large  community,  there  is  probably 
an  enameling  concern  which  can  wrinkle  your 
work  for  you  at  a reasonable  cost.  A very  at- 
tractive finish,  for  panels  especially,  is  to 
spray  a wrinkle  finish  with  aluminum  paint. 
In  any  painting  operation  (or  plating,  either, 
for  that  matter),  the  work  should  be  very 
thoroughly  cleaned  of  all  greases  and  oils. 

To  protect  brass  from  tarnish,  thoroughly 
cleanse  and  remove  the  last  trace  of  grease  by 
the  use  of  potash  and  water.  The  brass  must 
be  carefully  rinsed  with  water  and  dried;  but 
in  doing  it,  care  must  be  taken  not  to  handle 
any  portion  with  the  bare  hands  or  anything 
else  that  is  greasy.  Then  lacquer. 

Winding  Coils  Coils  are  of  two  general  types, 
those  using  a form  and  "air- 
wound”  types.  Neither  type  offers  any  particu- 
lar constructional  difficulties.  Figure  10  il- 
lustrates the  procedure  used  in  form  winding 
a coil.  If  the  winding  is  to  be  spaced,  the 
spacing  can  be  done  either  by  eye  or  a string 
or  another  piece  of  wire  may  be  wound  simul- 
taneously with  the  coil  wire  and  removed  after 
the  winding  is  in  place.  The  usual  procedure 
is  to  clamp  one  end  of  the  wire  in  a vise,  at- 
taching the  other  end  to  the  coil  form  and  with 
the  coil  form  in  hand,  walk  slowly  towards  the 
vise  winding  the  wire  but  at  the  same  time 
keeping  a strong  tension  on  the  wire  as  the 
form  is  rotated.  After  the  coil  is  wound,  if 
there  is  any  possibility  of  the  turns  slipping, 
the  completed  coil  is  either  entirely  coated 
with  a coil  or  Duco  cement  or  cemented  in 
those  spots  where  slippage  might  occur. 


Figure  12 
GOOD  SHOP 
LAYOUT  AIDS 
CAREFUL 
WORKMANSHIP 

Built  in  a corner  of  a 
garage,  this  shop  has  all 
features  necessary  for 
electronic  work.  Test  in- 
struments are  arranged 
on  shelves  above  bench. 
Numerous  outlets  reduce 
"haywire'*  produced  by 
tangled  line  cords.  Not 
shown  in  picture  ore  drill 
press  ana  sander  at  end 
of  left  bench 


V-h-f  and  u-h-f  coils  ate  commonly  wound 
of  heavy  enameled  wire  on  a form  and  then 
removed  from  the  form  as  in  figure  11.  If 
the  coil  is  long  or  has  a tendency  to  buckle, 
strips  of  polystyrene  or  a similar  material  may 
be  cemented  longitudinally  inside  the  coil.  Due 
allowance  must  be  made  for  the  coil  springing 
out  when  removed  from  the  form,  when  select- 
ing the  diameter  of  the  form. 

On  air  wound  coils  of  this  type,  spacing  be- 
tween turns  is  accomplished  after  removal 
from  the  form,  by  running  a pencil,  the  shank 
of  a screwdriver  or  other  round  object  spirally 
between  the  turns  from  one  end  of  the  coil  to 
the  other,  again  making  due  allowance  for 
spring. 

Air-wound  coils,  approaching  the  appearance 
of  commercially  manufactured  ones,  can  be 
constructed  by  using  a round  wooden  form 
which  has  been  sawed  diagonally  from  end 
to  end.  Strips  of  insulating  material  are  tem- 
porarily attached  to  this  mandrel,  the  wire 
then  being  wound  over  these  strips  with  the 
desired  separation  between  turns  and  cemented 
to  the  strips.  When  dry,  the  split  mandrel  may 
be  removed  by  unwedging  it. 

33-7  Shop  Layout 

The  size  of  your  workshop  is  relatively  un- 
important since  the  shop  layout  will  deter- 
mine its  efficiency  and  the  ease  with  which 
you  may  complete  your  work. 

Shown  in  figure  12  is  a workshop  built 
into  a lO’xlO’  area  in  the  corner  of  a garage. 
The  workbench  is  32”  wide,  made  up  of  four 
strips  of  2”x8”  lumber  supported  on  a solid 


framework  made  of  2”x4”  lumber.  The  top 
of  the  workbench  is  covered  with  hard-sur- 
face Masonite.  The  edge  of  the  surface  is  pro- 
tected with  aluminum  "counter  edging”  strip, 
obtainable  at  large  hardware  stores.  Two  wood- 
en shelves  12”  wide  are  placed  above  the 
bench  to  hold  the  various  items  of  test  equip- 
ment. The  shelves  are  bolted  to  the  wall  studs 
with  large  angle  brackets  and  have  wooden 
end  pieces.  Along  the  edge  of  the  lower  shelf 
a metal  "outlet  strip”  is  placed  that  has  an 
115-volt  outlet  every  six  inches  along  its 
length.  A similar  strip  is  run  along  the  back 
of  the  lower  shelf.  The  front  strip  is  used 
for  equipment  that  is  being  bench-tested,  and 
the  rear  strip  powers  the  various  items  of  test 
equipment  placed  on  the  shelves. 

At  the  left  of  the  bench  is  a storage  bin 
for  small  components.  A file  cabinet  can  be 
seen  at  the  right  of  the  photograph.  This  nec- 
essary item  holds  schematics,  transformer  data 
sheets,  and  other  papers  that  normally  are 
lost  in  the  usual  clutter  and  confusion. 

The  area  below  the  workbench  has  two 
storage  shelves  which  are  concealed  by  sliding 
doors  made  of  V4-inch  Masonite.  Heavier  tools, 
and  large  components  are  stored  in  this  area. 
On  the  floor  and  not  shown  in  the  photo- 
graph is  a very  necessary  item  of  shop  equip- 
ment: a large  trash  receptacle. 

A compact  and  efficient  shop  built  in  one- 
half  of  a wardrobe  closet  is  shown  in  figure 
13.  The  workbench  length  is  four  feet.  The 

top  is  made  of  4”x6’’  lumber  sheathed  with 

hard  surface  Masonite  and  trimmed  with 
"counter  edging”  strip.  The  supporting  struc- 


729-A 


730-A  Workshop  Practice 


Sslgife 


m 

K 

. m . = 

'it-. 

IK'  iKtTii 

."A#! 

« 

Figure  13 

COMPLETE  WORKSHOP  IN  A CLOSET! 

Careful  layout  permits  complete  electronic 
workshop  to  be  placed  in  one-half  of  a ward- 
robe closet.  Work  bench  is  built  atop  an  un- 
painted three-shelf  bookcase. 


ture  is  made  from  an  unpainted  three-shelf 
bookcase.  A 2”x2”  leg  is  placed  under  the 
front  corners  of  the  bench  to  provide  maxi- 
mum stability. 

Atop  the  bench,  a small  wooden  framework 
supports  needed  items  of  test  equipment  and 


a single  shelf  contains  a 1 15-volt  "outlet  strip.” 
The  instruments  at  the  top  of  the  photo  are 
placed  on  the  wardrobe  shelf. 

When  not  in  use,  the  doors  of  the  ward- 
robe are  closed,  concealing  the  workshop  com- 
pletely from  view. 


CHAPTER  THIRTY-FOUR 


Electronic 

Test  Equipment 


All  amateur  stations  are  required  by  kw  to 
have  certain  items  of  test  equipment  available 
within  the  station.  A c-w  station  is  required 
to  have  a frequency  meter  or  other  means  in 
addition  to  the  transmitter  frequency  control 
for  insuring  that  the  transmitted  signal  is  on 
a frequency  within  one  of  the  frequency  bands 
assigned  for  such  use.  A radiophone  station  is 
required  in  addition  to  have  a means  of  deter- 
mining that  the  transmitter  is  not  being  modu- 
lated in  excess  of  its  modulation  capability, 
and  in  any  event  not  mote  than  100  per  cent. 
Further,  any  station  operating  with  a power 
input  greater  than  900  watts  is  required  to 
have  a means  of  determining  the  exact  input 
to  the  final  stage  of  the  transmitter,  so  as  to 
insure  that  the  power  input  to  the  plate  circuit 
of  the  output  stage  does  not  exceed  1000 
watts. 

The  additional  test  and  measurement  equip- 
ment required  by  a station  will  be  determined 
by  the  type  of  operation  contemplated.  It  is 
desirable  that  all  stations  have  an  accurately 
calibrated  volt-ohmmeter  for  routine  transmit- 
ter and  receiver  checking  and  as  an  assistance 
in  getting  new  pieces  of  equipment  into  opera- 
tion. An  oscilloscope  and  an  audio  oscillator 
make  a very  desirable  adjunct  to  a phone  sta- 
tion using  AM  or  FM  transmission,  and  are  a 
necessity  if  single-sideband  operation  is  con- 
templated. A calibrated  signal  generator  is  al- 
most a necessity  if  much  receiver  work  is  con- 
templated, although  a frequency  meter  of  LM 
01  BC-221  type,  particularly  if  it  includes  in- 
ternal modulation,  will  serve  in  place  of  the 
signal  generator.  Extensive  antenna  work  in- 


variably requires  the  use  of  some  type  of  field- 
strength  meter,  and  a standing-wave  meter  of 
some  type  is  very  helpful.  Lastly,  if  much  v-h-f 
work  is  to  be  done,  a simple  grid-dip  meter 
will  be  found  to  be  one  of  the  most  used  items 
of  test  equipment  in  the  station. 

34-1  Voltage, 

Current  and  Power 

The  measurement  of  voltage  and  current  in 
radio  circuits  is  very  important  in  proper  main- 
tenance of  equipment.  Vacuum  tubes  of  the 
types  used  in  communications  work  must  be 
operated  within  rather  narrow  limits  in  regard 
to  filament  or  heater  voltage,  and  they  must 
be  operated  within  certain  maximum  limits  in 
regard  to  the  voltage  and  current  on  other 
electrodes. 

Both  direct  current  and  voltage  ate  most 
commonly  measured  with  the  aid  of  an  instru- 
ment consisting  of  a coil  that  is  free  to  rotate 
in  a constant  magnetic  field  {d’Arsonval  type 
instrument).  If  the  instrument  is  to  be  used  for 
the  measurement  of  current  it  is  called  an  am- 
meter or  milliammeter.  The  current  flowing 
through  the  circuit  is  caused  to  flow  through 
the  moving  coil  of  this  type  of  instrument.  If 
the  current  to  be  measured  is  greater  than  10 
milliamperes  or  so  it  is  the  usual  practice  to 
cause  the  majority  of  the  current  to  flow 
through  a by-pass  resistor  called  a shunt,  only 
a specified  portion  of  the  current  flowing 
through  the  moving  coil  of  the  instrument. 
The  calculation  of  shunts  for  extending  the 
range  of  d-c  milliammeters  and  ammeters  is 
discussed  in  Chapter  Two. 


721 -B 


722-B  Test  Equipment 


THE  RADIO 


+JO  -f-ioo  +250  +t000 


+ 10  +100  +250  +1000 


© 

Figure  1 


MULTI-VOLTMETER  CIRCUITS 

(A)  shows  a circuit  whereby  individual  multi- 
plier resistors  are  used  for  each  range.  (B)  is 
the  more  economical  ''series  multiplier"  cir- 
cuit, The  same  number  of  resistors  is  re- 
quired, but  those  for  the  higher  ranges  have 
less  resistance,  and  hence  are  less  expen- 
sive when  precision  wirewound  resistors  are 
to  be  used.  (C)  shows  a circuit  essentially 
the  same  as  at  (A),  except  that  a range 
switch  is  used.  With  a 0-500  d-c  microam- 
meter substituted  for  the  0- 1 milUammeter 
shown  above,  ali  resistor  values  would  be 
multiplied  by  two  and  the  voltmeter  would 
have  a "2000-ohm-per-volt"  sensitivity.  Simi- 
larly, if  a O-SO  d-c  microammeter  were  to  be 
used,  all  resistance  values  would  be  multi- 
plied by  twenty,  and  the  voltmeter  would 
have  a sensitivity  of  20,000  ohms  per  volt. 


A direct  current  voltmeter  is  merely  a d-c 
milliammeter  with  a multiplier  resistor  in  se- 
ries with  it.  If  it  is  desired  to  use  a low-range 
milliammeter  as  a voltmeter  the  value  of  the 
multiplier  resistor  for  any  voltage  range  may 
be  determined  from  the  following  formula; 

1000  E 

R = 

I 

where;  R = multiplier  resistor  in  ohms 
E = desired  full  scale  voltage 
I = full  scale  current  of  meter  in  ma. 

The  sensitivity  of  a voltmeter  is  commonly 
expressed  in  ohms  per  volt.  The  higher  the 
ohms  per  volt  of  a voltmeter  the  greater  its 
sensitivity.  When  the  full-scale  current  drain 
of  a voltmeter  is  known,  its  sensitivity  rating 
in  ohms  per  volt  may  be  determined  by; 

1000 

Ohms  per  volt  = 

I 


With  the  switch  In  position  1,  the  0-7  milliam- 
meier  would  be  connected  directly  to  the 
terminals.  In  position  2 the  meter  would  read 
from  0-100,000  ohms,  approximately,  with  a 
resistance  value  of  4500  ohms  at  half  scale. 
(Note:  The  half-scale  resistance  value  of  an 
ohmmeter  using  this  circuit  is  equal  to  the 
resistance  in  series  with  the  battery  inside 
the  instrument.)  The  other  four  taps  are  volt- 
age ranges  with  10,  SO,  250,  and  500  volts 
full  5co/e. 


Where  I is  the  full-scale  current  drain  of  the 
indicating  instrument  in  milliamperes. 

Multi-Range  It  is  common  practice  to  connect 
Meters  a group  of  multiplier  resistors 

in  the  circuit  with  a single  in- 
dicating instrument  to  obtain  a multi-range 
voltmeter.  There  are  several  ways  of  wiring 
such  a meter,  the  most  common  ones  of  which 
are  indicated  in  figure  1.  With  all  these  meth- 
ods of  connection,  the  sensitivity  of  the  meter 
in  ohms  per  volt  is  the  same  on  all  scales. 
With  a 0-1  milliammeter  as  shown  the  sensi- 
tivity is  1000  ohms  per  volt. 

Volf-Ohmmeters  An  extremely  useful  piece  of 
test  equipment  which  should 
be  found  in  every  laboratory  or  radio  station 
is  the  volt-ohmmeter.  It  consists  of  a multi- 
range voltmeter  with  an  additional  fixed  resis- 
tor, a variable  resistor,  and  a battery.  A typical 
example  of  such  an  instrument  is  diagrammed 
in  figure  2.  Tap  1 is  used  to  permit  use  of 
the  instrument  as  an  0-1  d-c  milliammeter. 
Tap  2 permits  accurate  reading  of  resistors  up 
to  100,000  ohms;  taps  3,  4,  5,  and  6 are 
for  making  voltage  measurements,  the  full 
scale  voltages  being  10,  50,  250,  and  500 
volts  respectively. 

The  1000-ohm  potentiometer  is  used  to 
bring  the  needle  to  zero  ohms  when  the  ter- 
minals are  shorted;  this  adjustment  should 
always  be  made  before  a resistance  measure- 
ment is  taken.  Higher  voltages  than  500  can 
be  read  if  a higher  value  of  multiplier  resistor 
is  added  to  an  additional  tap  on  the  switch. 
The  proper  value  for  a given  full  scale  read- 
ing can  be  determined  from  Ohm’s  law. 

Resistances  higher  than  100,000  ohms  can- 


HANDBOOK 


Ohmmeters  723-B 


not  be  measured  accurately  with  the  circuit 
constants  shown;  however,  by  increasing  the 
ohmmeter  battery  to  45  volts  and  multiplying 
the  4000-ohm  resistor  and  1000-ohm  poten- 
tiometer by  10,  the  ohms  scale  also  will  be 
multiplied  by  10.  This  would  permit  accurate 
measurements  up  to  1 megohm. 

0-1  d-c  milliammeters  are  available  with 
special  volt-ohmmeter  scales  which  make  in- 
dividual calibration  unnecessary.  Or,  special 
scales  can  be  purchased  separately  and  sub- 
stituted for  the  original  scale  on  the  milliam- 
meter. 

Obviously,  the  accuracy  of  the  instrument 
either  as  a voltmeter  or  as  an  ammeter  can  be 
no  better  than  the  accuracy  of  the  milliam- 
meter  and  the  resistors. 

Because  volt-ohmmeters  are  so  widely  used 
and  because  the  circuit  is  standardized  to  a 
considerable  extent,  it  is  possible  to  purchase 
a factory-built  volt-ohmmeter  for  no  more  than 
the  component  parts  would  cost  if  purchased 
individually.  For  this  reason  no  construction 
details  are  given.  However,  anyone  already 
possessing  a suitable  milliammeter  and  de- 
sirous of  incorporating  it  in  a simple  volt- 
ohmmeter  should  be  able  to  build  one  from  the 
schematic  diagram  and  design  data  given  here. 
Special,  precision  (accurately  calibrated)  mul- 
tiplier resistors  are  available  if  a high  degree 
of  accuracy  is  desired.  Alternatively,  good 
quality  carbon  resistors  whose  actual  resist- 
ance has  been  checked  may  be  used  as  multi- 
pliers where  less  accuracy  is  required. 

Medium-  and  Most  ohmmeters,  including  the 
Low-Range  one  just  described,  are  not 

Obrnmeter  adapted  for  accurate  measure- 

ment of  low-resistances — in  the 
neighborhood  of  100  ohms,  for  instance. 

The  ohmmeter  diagrammed  in  figure  3 was 
especially  designed  for  the  reasonably  accu- 
rate reading  of  resistances  down  to  1 ohm.  Two 
scales  are  provided,  one  going  in  one  direc- 
tion and  the  other  scale  going  in  the  other 
direction  because  of  the  different  manner  in 
which  the  milliammeter  is  used  in  each  case. 
The  low  scale  covers  from  1 to  100  ohms  and 
the  high  scale  from  100  to  10,000  ohms.  The 
high  scale  is  in  reality  a medium-range  scale. 
For  accurate  reading  of  resistances  over  10,000 
ohms,  an  ohmmeter  of  the  type  previously 
described  should  be  used. 

The  1-100  ohm  scale  is  useful  for  checking 
transformers,  chokes,  r-f  coils,  etc.,  which  often 
have  a resistance  of  only  a few  ohms. 

The  calibration  scale  will  depend  upon  the 
internal  resistance  of  the  particular  make  of 


D.  P,  D.T. 
SWITCH 


OHMMETER 

A description  of  the  operation  of  this  circuit 
is  given  in  the  text.  With  the  switch  in  the 
left  position  the  half-scale  reading  of  the 
meter  will  occur  with  an  external  resistance 
of  1000  ohms.  With  the  switch  in  the  right 
position,  half-scale  deflection  will  be  ob- 
tained with  an  external  resistance  equal  to 
the  d-c  resistance  of  the  milliammeter  (20  to 
50  ohms  depending  upon  the  make  of  in- 
strument). 


1.5-ma.  meter  used.  The  instrument  can  be 
calibrated  by  means  of  a Wheatstone  bridge  or 
a few  resistors  of  known  accuracy.  The  latter 
can  be  series-connected  and  parallel-connected 
to  give  sufficient  calibration  points.  A hand- 
drawn  hand-calibrated  scale  can  be  cemented 
over  the  regular  meter  scale  to  give  a direct 
reading  in  ohms. 

Before  calibrating  the  instrument  or  using 
it,  the  test  prods  should  always  be  touched  to- 
gether and  the  zero  adjuster  set  accurately. 

Measurement  of  The  measurement  of  al- 
Alternoting  Current  ternating  current  and 
and  Voltoge  voltage  is  complicated 

by  two  factors;  first,  the 
frequency  range  covered  in  ordinary  communi- 
cation channels  is  so  great  that  calibration  of 
an  instrument  becomes  extremely  difficult;  sec- 
ond, there  is  no  single  type  of  instrument 
which  is  suitable  for  all  a-c  measurements — as 
the  d’Arsonval  type  of  movement  is  suitable 
for  d-c.  The  d’Arsonval  movement  will  not 
operate  on  a-c  since  it  indicates  the  average 
value  of  current  flow,  and  the  average  value 
of  an  a-c  wave  is  zero. 

As  a result  of  the  inability  of  the  reliable 
d’Arsonval  type  of  movement  to  record  an  al- 
ternating current,  either  this  current  must  be 
rectified  and  then  fed  to  the  movement,  or  a 
special  type  of  movement  which  will  operate 
from  the  effective  value  of  the  current  can  be 
used. 

For  the  usual  measurements  of  power  fre- 
quency a.c.  (25-60  cycles)  the  Iron-vane  in- 
strument is  commonly  used.  For  audio  fre- 
quency a.c.  (50-20,000  cycles)  a d’Arsonval 


724-B  Test  Equipment 


THE  RADIO 


1G4-G 


Figure  4 

SLIDE-BACK  V-T  VOLTMETER 

By  connecting  a variable  source  of  voltage 
in  series  with  the  input  to  a conventional 
v-t  voltmeter,  or  in  series  with  the  simple 
trlode  voltmeter  shown  above,  a slide-back 
a-c  voltmeter  for  peak  voltage  measurement 
con  be  constructed.  Resistor  ft  should  be 
about  1000  ohms  per  volt  used  at  battery  B. 
This  type  of  v-t  voltmeter  has  the  advantage 
that  it  can  give  a reading  of  the  actual  peak 
voltage  of  the  wave  being  measured,  without 
any  current  drain  from  the  source  of  voltage. 


instrument  having  an  integral  copper  oxide 
or  selenium  rectifier  is  usually  used.  Radio  fre- 
quency voltage  measurements  are  usually  made 
with  some  type  of  vacuum-tube  voltmeter, 
while  r-f  current  measurements  are  almost  in- 
variably made  with  an  instrument  containing 
a thermo-couple  to  convert  the  r.f.  into  d.c. 
for  the  movement. 

Since  an  alternating  current  wave  can  have 
an  almost  infinite  variety  of  shapes,  it  can 
easily  be  seen  that  the  ratios  between  the 
three  fundamental  quantities  of  the  wave 
(peak,  r.m.s.  effective,  and  average  after  rec- 
tification) can  also  vary  widely.  So  it  becomes 
necessary  to  know  beforehand  just  which  qual- 
ity of  the  wave  under  measurement  our  in- 
strument is  going  to  indicate.  For  the  purpose 
of  simplicity  we  can  list  the  usual  types  of 
a-c  meters  in  a table  along  with  the  charac- 
teristic of  an  a-c  wave  which  they  will  indi- 
cate: 

Iron-vane,  thermocouple — r.m.s. 

Rectifier  type  (copper  oxide  or  selenium) 
— average  after  rectification. 

V.t.v.m. — r.m.s.,  average  or  peak,  depend- 
ing upon  design  and  calibration. 

Vacuum-Tube  A vacuum-tube  voltmeter  is  es- 
Voltmeters  sentially  a detector  in  which 
a change  in  the  signal  placed 
upon  the  input  will  produce  a change  in  the 
indicating  instrument  (usually  a d’Atsonval 
meter)  placed  in  the  output  circuit.  A vacuum- 
tube  voltmeter  may  use  a diode,  a triode,  or  a 
multi-element  tube,  and  it  may  be  used  either 
for  the  measurement  of  a.c.  or  d.c. 

When  a v.t.v.m.  is  used  in  d-c  measurement 
it  is  used  for  this  purpose  primarily  because 
of  the  very  great  input  resistance  of  the  device. 


This  means  that  a v.t.v.m.  may  be  used  for 
the  measurement  of  a-v-c,  a-f-c,  and  discrimina- 
tor output  voltages  where  no  loading  of  the 
circuit  can  be  tolerated. 

A-C  V-T  There  are  many  different  types 
Voltmeters  of  a-c  vacuum-tube  voltmeters, 
all  of  which  operate  as  some 
type  of  rectifier  to  give  an  indication  on  a d-c 
instrument.  There  are  two  general  types:  those 
which  give  an  indication  of  the  t-m-s  value  of 
the  wave  (or  approximately  this  value  of  a 
complex  wave),  and  those  which  give  an  in- 
dication of  the  peak  or  crest  value  of  the 
wave. 

Since  the  setting  up  and  calibration  of  a 
wide-range  vacuum-tube  voltmeter  is  rather 
tedious,  in  most  cases  it  will  be  best  to  pur- 
chase a commercially  manufactured  unit.  Sev- 
eral excellent  commercial  units  are  on  the 
market  at  the  present  time;  also  kits  for  home 
construction  of  a quite  satisfactory  v.t.v.m. 
are  available  from  several  manufacturers. 
These  feature  a wide  range  of  a-c  and  d-c  volt- 
age scales  at  high  sensitivity,  and  in  addition 
several  feature  a built-in  vacuum-tube  ohm- 
meter  which  will  give  indications  up  to  500 
or  1000  megohms. 

Peak  A-C  V-T  There  are  two  common  types 
Voltmeters  of  peak-indicating  vacuum- 
tube  voltmeters.  The  first  is 
the  so-called  slide-back  type  in  which  a sim- 
ple v.t.v.m.  is  used  along  with  a conventional 
d-c  voltmeter  and  a source  of  bucking  bias  in 
series  with  the  input.  With  this  type  of  ar- 
rangement (figure  4)  leads  are  connected  to 
the  voltage  to  be  measured  and  the  slider  re- 
sistor R across  the  bucking  voltage  is  backed 
down  until  an  indication  on  the  meter  (called 
a false  zero)  equal  to  that  value  given  with 
the  prods  shorted  and  the  bucking  voltage 
reduced  to  zero,  is  obtained.  Then  the  value 
of  the  bucking  voltage  (read  on  V)  is  equal 
to  the  peak  value  of  the  voltage  under  meas- 
urement. The  slide-back  voltmeter  has  the 
disadvantage  that  it  is  not  instantaneous  in  its 
indication — adjustments  must  be  made  for 
every  voltage  measurement.  For  this  reason  the 
slide-back  v.t.v.m.  is  not  commonly  used,  be- 
ing supplanted  by  the  diode-rectifier  type  of 
peak  v.t.v.m.  for  most  applications. 

High-Voltage  A diode  vacuum-tube  voltmeter 
Diode  Peek  suitable  for  the  measurement 
Voltmeter  of  high  values  of  a-c  voltage  is 
diagrammed  in  figure  5.  With 
the  constants  shown,  the  voltmeter  has  two 


HANDBOOK 


Power  Measurement  725-B 


2X2/879 


A.C. 

VOLTAGE 

FROM 

SOURCE 

WITH 

D.C. 

RETURN 

PATH 


A.C. 

VOLTAGE 

FROM 

SOURCE 

WITH 

O.C. 

RETURN 

PATH 


CONVENTIONAL 

HIGH-SENSITIVITY 

VOLTMETER 


Figure  5 

SCHEMATIC  OF  A HIGH-VOLTAGE  PEAK 
VOLTMETER 

A peak  voltmeter  such  as  diagrammed 
above  is  convenient  for  the  measure- 
ment of  peak  voltages  at  fairly  high 
power  levels  from  a source  of  moder* 
ately  low  impedance. 

Cl — .OOl-jifd-  high-voltage  mica 
Ct — 7.0/x/d.  high-voltage  paper 
fli — 500,000  ohms  (two  0.2S-megohm 
in  series.) 

Ri — 1.0  megohm  (four  0.2S-megohm  5^-woff 
in  series} 

T — 2.5  V.,  1.75  a.  filament  transformer 
M — 0-1  d-c  milliammeter 

— ^~P-d-t  toggle  switch 
5 — S-p-s-t  toggle  switch 

(Note:  Cl  is  a by-pass  around  Ct,  the  induc- 
tive reocfance  of  which  may  be  appreciable 
at  high  frequencies.) 


ranges:  500  and  1500  volts  peai  full  scale. 

Capacitors  Ci  and  O should  be  able  to  with- 
stand a voltage  in  excess  of  the  highest  peak 
voltage  to  be  measured.  Likewise,  Ri  and  Rs 
should  be  able  to  withstand  the  same  amount 
of  voltage.  The  easiest  and  least  expensive 
way  of  obtaining  such  resistors  is  to  use  sev- 
eral low- voltage  resistors  in  series,  as  shown 
in  figure  5.  Other  voltage  ranges  can  be  ob- 
tained by  changing  the  value  of  these  re- 
sistors, but  for  voltages  less  than  several  hun- 
dred volts  a more  linear  calibration  can  be  ob- 
tained by  using  a receiving-type  diode.  A cali- 
bration curve  should  be  run  to  eliminate  the 
appreciable  error  due  to  the  high  internal  re- 
sistance of  the  diode,  preventing  the  capacitor 
from  charging  to  the  full  peak  value  of  the 
voltage  being  measured. 

A direct  reading  diode  peak  voltmeter  of 
the  type  shown  in  figure  5 will  load  the  source 
of  voltage  by  approximately  one-half  the  value 
of  the  load  resistance  in  the  circuit  (Ri,  or  Ri 
plus  Ri,  in  this  case).  Also,  the  peak  voltage 
reading  on  the  meter  will  be  slightly  less  than 
the  actual  peak  voltage  being  measured.  The 
amount  of  lowering  of  the  reading  is  deter- 
mined by  the  ratio  of  the  reactance  of  the 
storage  capacitance  to  the  load  resistance.  If 
a cathode-ray  oscilloscope  is  placed  across  the 
terminals  of  the  v.t.v.m.  when  a voltage  is  be- 
ing measured,  the  actual  amount  of  the  lower- 
ing in  voltage  may  be  determined  by  inspection 


Figure  6 

PEAK- VOLTAGE  MEASUREMENT 
CIRCUIT 

Through  use  of  the  arrangement  shown  above 
it  is  possible  to  make  accurate  measurements 
of  peak  a-e  voltages,  such  as  across  the  sec- 
ondary of  a modulation  frans/ormer,  with  a 
conventional  d-c  multi-voltmeter.  Capacitor 
C and  transformer  T should,  of  course,  be  in- 
sulated for  the  highest  peak  voltage  likely  to 
be  encountered.  A capacitance  of  0.25-/xfd. 
at  C has  been  tound  to  be  adequate.  The 
higher  the  sensitivity  of  the  indicating  d-c 
voltmeter,  the  smaller  will  be  the  error  be- 
tween the  indication  on  the  meter  and  the  ac- 
tual peak  voltage  being  measured. 


of  the  trace  on  the  c-r  tube  screen.  The  peak 
positive  excursion  of  the  wave  will  be  slightly 
flattened  by  the  action  of  the  v.t.v.m.  Usually 
this  flattening  will  be  so  small  as  to  be  negli- 
gible. 

An  alternative  arrangement,  shown  in  figure 
6,  is  quite  convenient  for  the  measurement  of 
high  a-c  voltages  such  as  are  encountered  in 
the  adjustment  and  testing  of  high-power 
audio  amplifiers  and  modulators.  The  arrange- 
ment consists  simply  of  a 2X2  rectifier  tube 
and  a filter  capacitor  of  perhaps  0.25-gfd.  ca- 
pacitance, but  with  a voltage  rating  high 
enough  that  it  is  not  likely  to  be  punaured 
as  a result  of  any  tests  made.  Cathode-ray 
oscilloscope  capacitors,  and  those  for  electro- 
static-deflection TV  tubes  often  have  ratings 
as  high  as  0.25  /xfd.  at  7500  to  10,000  volts. 
The  indicating  instrument  is  a conventional 
multi-scale  d-c  voltmeter  of  the  high-sensitivity 
type,  preferably  with  a sensitivity  of  20,000  or 
50,000  ohms  per  volt.  The  higher  the  sen- 
sitivity of  the  d-c  voltmeter  used  with  the 
rectifier,  the  smaller  will  be  the  amount  of 
flattening  of  the  a-c  wave  as  a result  of  the 
rectifier  action. 

Measurement  Audio  frequency  or  radio  fre- 
of  Power  quency  power  in  a resistive 

circuit  is  most  commonly  and 
most  easily  determined  by  the  indirect  method, 
i.e.,  through  the  use  of  one  of  the  following 
formulas : 

P = El  P ^ EVR  P = PR 
These  three  formulas  mean  that  if  any  two  of 
the  three  factors  determining  power  are  known 


726-B  Test  Equipment 


TH  E RADIO 


Figure  7 

100-WATT  DUMMY  LOAD  SUITABLE  FOR 
RANGER,  OR  SIMILAR  TYPE 
TRANSMITTER 


(resistance,  current,  voltage)  the  power  being 
dissipated  may  be  determined.  In  an  ordinary 
120-volt  a-c  line  circuit  the  above  formulas 
are  not  strictly  true  since  the  power  factor  of 
the  load  must  be  multiplied  into  the  result — 
or  a direct  method  of  determining  power  such 
as  a wattmeter  may  be  used.  But  in  a resistive 
a-f  circuit  and  in  a resonant  r-f  circuit  the 
power  factor  of  the  load  is  taken  as  being 
unity. 

For  accurate  measurement  of  a-f  and  r-f 
power,  a thermogalvanometer  or  thermocouple 
ammeter  in  series  with  a non-inductive  resis- 
tor of  known  resistance  can  be  used.  The  me- 
ter should  have  good  accuracy,  and  the  exact 
value  of  resistance  should  be  known  with  ac- 
curacy. Suitable  dummy  load  resistors  are 
available  in  various  resistances  in  both  100 
and  250-watt  ratings.  These  are  virtually  non- 
inductive,  and  may  be  considered  as  a pure 
resistance  up  to  30  Me.  The  resistance  of 
these  units  is  substantially  constant  for  all 
values  of  current  up  to  the  maximum  dissipa- 
tion rating,  but  where  extreme  accuracy  is  re- 
quired, a correction  chart  of  the  dissipation 
coefficient  of  resistance  ( supplied  by  the  manu- 
facturer) may  be  employed.  This  chart  shows 


the  exact  resistance  for  different  values  of 
current  through  the  resistor. 

Sine- wave  power  measurements  (r-f  or  sin- 
gle-frequency audio)  may  also  be  made 
through  the  use  of  a v.t.v.m.  and  a resistor  of 
known  value.  In  fact  a v.t.v.m.  of  the  type 
shown  in  figure  6 is  particularly  suited  to  this 
work.  The  formula,  P = EVR  is  used  in  this 
case.  However,  it  must  be  remembered  that 
a v.t.v.m.  of  the  type  shown  in  figure  6 indi- 
cates the  peak  value  of  the  a-c  wave.  This 
reading  must  be  converted  to  the  r-m-s  or 
heating  value  of  the  wave  by  multiplying  it 
by  0.707  before  substituting  the  voltage  value 
in  the  formula.  The  same  result  can  be  ob- 
tained by  using  the  formula  P = EV2R. 

Thus  all  three  methods  of  determining  pow- 
er, ammeter-resistor,  voltmeter-resistor,  and 
voltmeter-ammeter,  give  an  excellent  cross- 
check upon  the  accuracy  of  the  determination 
and  upon  the  accuracy  of  the  standards. 

Power  may  also  be  measured  through  the 
use  of  a calorimeter,  by  actually  measuring  the 
amount  of  heat  being  dissipated.  Through  the 
use  of  a water-cooled  dummy  load  resistor  this 
method  of  power  output  determination  is  being 
used  by  some  of  the  most  modern  broadcast 
stations.  But  the  method  is  too  cumbersome 
for  ordinary  power  determinations. 

Power  may  also  be  determined  photometri- 
cally through  the  use  of  a voltmeter,  ammeter, 
incandescent  lamp  used  as  a load  resistor,  and 
a photographic  exposure  meter.  With  this 
method  the  exposure  meter  is  used  to  deter- 
mine the  relative  visual  output  of  the  lamp 
running  as  a dummy  load  resistor  and  of  the 
lamp  running  from  the  120-volt  a-c  line.  A 
rheostat  in  series  with  the  lead  from  the  a-c 
line  to  the  lamp  is  used  to  vary  its  light  inten- 
sity to  the  same  value  (as  indicated  by  the  ex- 
posure meter)  as  it  was  putting  out  as  a dum- 
my load.  The  a-c  voltmeter  in  parallel  with  the 
lamp  and  ammeter  in  series  with  it  is  then 
used  to  determine  lamp  power  input  by: 
P = El.  This  method  of  power  determination 
is  satisfactory  for  audio  and  low  frequency 
r.f.  but  is  not  satisfactory  for  v-h-f  work  be- 
cause of  variations  in  lamp  efficiency  due  to 
uneven  heating  of  the  filament. 

Dummy  Loads  Lamp  bulbs  make  poor  dummy 
loads  for  r-f  work,  in  general, 
as  they  have  considerable  reactance  above  2 
Me.,  and  the  resistance  of  the  lamp  varies  with 
the  amount  of  current  passing  through  it. 

A suitable  r-f  load  for  powers  up  to  a few 
watts  may  be  made  by  paralleling  2-watt  com- 
position resistors  of  suitable  value  to  make 


HANDBOOK 


Measurement  of  Constants  727-B 


a 50-ohm  resistor  of  5 or  10  watts  dissipation. 
The  resistors  should  be  mounted  in  a compact 
bundle,  with  short  interconnecting  leads. 

For  powers  of  100  and  250  watts,  the 
Ohmite  models  D-101  and  D-251  low  in- 
ductance resistors  may  be  used  as  dummy  an- 
tennas. A D-101  resistor  mounted  on  a small 
metal  chassis  with  a r-f  ammeter  and  coaxial 
plug  is  shown  in  figure  7.  The  Ohmite  D-250 
resistor  may  be  employed  in  such  a configura- 
tion for  higher  power. 

A dummy  load  capable  of  taking  the  output 
of  a plate-modulated  1 -kilowatt  transmitter  is 
shown  in  figure  8.  The  load  is  made  up  of  ten 
160-watt,  500-ohm  Ohmite  type  2409  non- 
inductive  resistors.  These  resistors  are  mounted 
between  two  aluminum  end  plates  measuring 
4 inches  in  diameter.  The  assembly  is  then 
end  mounted  by  means  of  1”  ceramic  insula- 
tors to  an  aluminum  chassis  containing  the 
r-f  ammeter  and  an  SO-239  coaxial  receptacle. 

This  dummy  load  presents  a good  approxi- 
mation of  a 50-ohm,  1600-watt  resistor  up 
to  approximately  8 Me.  Above  this  frequency, 
it  becomes  increasingly  reactive.  It  may  be 
used  up  to  30  Me.,  if  it  is  remembered  that 
the  reading  of  the  r-f  ammeter  no  longer  is 
an  absolute  indication  of  the  resistive  power 
dissipation  of  the  load.  For  use  on  the  20, 
15  and  10  meter  bands,  it  is  recommended 
that  the  coaxial  line  coupling  the  dummy  load 
to  the  transmitter  either  be  restriaed  to  a 
length  of  two  feet  or  less,  or  else  be  made 
an  electrical  half-wavelength  long.  This  will 
insure  that  a minimum  of  reactance  will  be 
coupled  back  into  the  transmitter. 

34-2  Measuremenf 

of  Circuit  Constants 

The  measurement  of  the  resistance,  capaci- 
tance, inductance,  and  Q (figure  of  merit)  of 
the  components  used  in  communications  work- 
can  be  divided  into  three  general  methods: 
the  impedance  method,  the  substitution  or  reso- 
nance method,  and  the  bridge  method. 

The  Impedance  The  impedance  method  of 

Method  measuring  inductance  and 

capacitance  can  be  likened  to 
the  ohmmeter  method  for  measuring  resistance. 
An  a-c  voltmeter,  or  milliammeter  in  series 
with  a resistor,  is  connected  in  series  with 
the  inductance  or  capacitance  to  be  measured 
and  the  a-c  line.  The  reading  of  the  meter  will 
be  inversely  proportional  to  the  impedance  of 
the  component  being  measured.  After  the  me- 
ter has  been  calibrated  it  will  be  possible  to 


R - TEN  500-OHM,  160-WATT  NON-INDUCTIVE  RESISTORS 
IN  PARALLEL  {OHMITE  2.409) 

Figure  8 

1.6-KILOWATT  DUMMY  LOAD  SCHEMATIC 

obtain  the  approximate  value  of  the  impedance 
directly  from  the  scale  of  the  meter.  If  the 
component  is  a capacitor,  the  value  of  im- 
pedance may  be  taken  as  its  reactance  at  the 
measurement  frequency  and  the  capacitance 
determined  accordingly.  But  the  d-c  resistance 
of  an  inductor  must  also  be  taken  into  con- 
sideration in  determining  its  inductance.  After 
the  d-c  resistance  and  the  impedance  have 
been  determined,  the  reactance  may  be  deter- 
mined from  the  formula;  Xl=VZ“~R“. 
Then  the  inductance  may  be  determined  from: 
L = XL/2irf. 

The  Substil-ution  The  substitution  method  is 
Method  a satisfactory  system  for 

obtaining  the  inductance  or 
capacitance  of  high-frequency  components.  A 
large  variable  capacitor  with  a good  dial  hav- 
ing an  accurate  calibration  curve  is  a neces- 
sity for  making  determinations  by  this  method. 
If  an  unknown  inductor  is  to  be  measured,  it  is 
connected  in  parallel  with  the  standard  capaci- 
tor and  the  combination  tuned  accurately  to 
some  known  frequency.  This  tuning  may  be 
accomplished  either  by  using  the  tuned  circuit 
as  a wavemeter  and  coupling  it  to  the  tuned 
circuit  of  a reference  oscillator,  or  by  using 
the  tuned  circuit  in  the  controlling  position  of 
a two  terminal  oscillator  such  as  a dynatron 
or  transitron.  The  capacitance  required  to  tune 
this  first  frequency  is  then  noted  as  Ci.  The 
circuit  or  the  oscillator  is  then  tuned  to  the 
second  harmonic  of  this  first  frequency  and 
the  amount  of  capacitance  again  noted,  this 

time  as  &.  Then  the  distributed  capacitance 

across  the  coil  (including  all  stray  capaci- 
tances) is  equal  to:  Co=  (Ci“4C2)/3. 


728-B  Test  Equipment 


THE  RADIO 


@ ® 

Figure  9 

TWO  WHEATSTONE  BRIDGE  CIRCUITS 

These  circuits  are  used  for  the  measuremertt 
of  d-c  resistance.  In  (A)  the  “ratio  arms“  ftg 
and  are  fixed  and  balancing  of  the  bridge 
is  accomplished  by  variation  of  the  standard 
R^.  The  standard  in  this  case  usually  con- 
sists of  a decade  box  giving  resistance  in 
l-ohm  steps  from  Q to  1 1 JO  or  to  J J,1 10  ohms. 
In  (B)  a fixed  standard  is  used  for  each  range 
end  the  ratio  arm  is  varied  to  obtain  balance, 
A calibrated  slide-wire  or  potentiometer  cali- 
brated by  resistance  in  terms  of  degrees  is 
usually  employed  as  R^  and  R^.  It  will  be 
noticed  that  the  formula  for  determining  the 
unknown  resistance  from  the  known  is  the 
same  in  either  case. 


This  value  of  distributed  capacitance  is 
then  substituted  in  the  following  formula  along 
with  the  value  of  the  standard  capacitance  for 
either  of  the  two  frequencies  of  measurement; 

1 

L = 

47rTi“(G  + Co) 

The  determination  of  an  unknown  capaci- 
tance is  somewhat  less  complicated  than  the 
above.  A tuned  circuit  including  a coil,  the 
unknown  capacitor  and  the  standard  capacitor, 
all  in  parallel,  is  resonated  to  some  conveni- 
ent frequency.  The  capacitance  of  the  stand- 
ard capacitor  is  noted.  Then  the  unknown  ca- 
pacitor is  removed  and  the  circuit  re-resonated 
by  means  of  the  standard  capacitor.  The  dif- 
ference between  the  two  readings  of  the  stand- 
ard capacitor  is  then  equal  to  the  capacitance 
of  the  unknown  capacitor. 

34-3  Measurements 
with  a Bridge 

Experience  has  shown  that  one  of  the  most 
satisfactory  methods  for  measuring  circuit  con- 
stants (resistance,  capacitance,  and  inductance) 
at  audio  frequencies  is  by  means  of  the  a-c 
bridge.  The  Wheatstone  {d-c)  bridge  is  also 
one  of  the  most  accurate  methods  for  the 
measurement  of  d-c  resistance.  With  a simple 
bridge  of  the  type  shown  at  figure  9A  it  is 
entirely  practical  to  obtain  d-c  resistance  de- 


terminations accurate  to  four  significant  fig- 
ures. With  an  a-c  bridge  operating  within  its 
normal  rating  as  to  frequency  and  range  of 
measurement  it  is  possible  to  obtain  results 
accurate  to  three  significant  figures. 

Both  the  a-c  and  the  d-c  bridges  consist  of 
a source  of  energy,  a standard  or  reference  of 
measurement,  a means  of  balancing  this  stand- 
ard against  the  unknown,  and  a means  of  in- 
dicating when  this  balance  has  been  reached. 
The  source  of  energy  in  the  d-c  bridge  is  a 
battery;  the  indicator  is  a sensitive  galvanome- 
ter. In  the  a-c  bridge  the  source  of  energy 
is  an  audio  oscillator  (usually  in  the  vicinity 
of  1000  cycles),  and  the  indicator  is  usually 
a pair  of  headphones.  The  standard  for  the  d-c 
bridge  is  a resistance,  usually  in  the  form  of 
a decade  box.  Standards  for  the  a-c  bridge  can 
be  resistance,  capacitance,  and  inductance  in 
varying  forms. 

Figure  9 shows  two  general  types  of  the 
Wheatstone  or  d-c  bridge.  In  ( A ) the  so-called 
"ratio  arms”  Ra  and  R®  are  fixed  (usually  in 
a ratio  of  1-to-l,  1-to-lO,  1 -to- 100,  or  1-to- 
1,000)  and  the  standard  resistor  Rs  is  varied 
until  the  bridge  is  in  balance.  In  commercially 
manufactured  bridges  there  are  usually  two  or 
more  buttons  on  the  galvanometer  for  progres- 
sively increasing  its  sensitivity  as  balance  is 
approached.  Figure  9B  is  the  slide  wire  type 
of  bridge  in  which  fixed  standards  are  used 
and  the  ratio  arm  is  continuously  variable. 
The  slide  wire  may  actually  consist  of  a mov- 
ing contact  along  a length  of  wire  of  uni- 
form cross  section  in  which  case  the  ratio  of 
Ra  to  Rb  may  be  read  off  direaly  in  centi- 
meters or  inches,  or  in  degrees  of  rotation  if 
the  slide  wire  is  bent  around  a circular  former. 
Alternatively,  the  slide  wire  may  consist  of 
linear-wound  potentiometer  with  its  dial  cali- 
brated in  degrees  or  in  resistance  from  each 
end. 

Figure  lOA  shows  a simple  type  of  a-c 
bridge  for  the  measurement  of  capacitance  and 
inductance.  It  can  also,  if  desired,  be  used 
for  the  measurement  of  resistance.  The  four 
arms  of  the  bridge  may  be  made  up  in  a va- 
riety of  ways.  As  before,  R®  and  Ra  make  up 
the  ratio  arms  of  the  device  and  may  be  either 
of  the  slide  wire  type,  as  indicated,  or  they 
may  be  fixed  and  a variable  standard  used  to 
obtain  balance.  In  any  case  it  is  always  neces- 
sary with  this  type  of  bridge  to  use  a standard 
which  presents  the  same  type  of  impedance  as 
the  unknown  being  measured;  resistance  stand- 
ard for  a resistance  measurement,  capacitance 
standard  for  capacitance,  and  inductance  stand- 


HANDBOOK 


Frequency  Measurement  729-B 


ard  for  inductance  determination.  Also,  k is  a 
great  help  in  obtaining  an  accurate  balance  of 
the  bridge  if  a standard  of  approximately  the 
same  value  as  the  assumed  value  of  the  un- 
known is  employed.  Also,  the  standard  should 
be  of  the  same  general  type  and  should  have 
approximately  the  same  power  factor  as  the  un- 
known impedance.  If  all  these  precautions  are 
observed,  little  trouble  will  be  experienced  in 
the  measurement  of  resistance  and  in  the 
measurement  of  impedances  of  the  values 
usually  used  in  audio  and  low  radio  frequency 
work. 

However,  the  bridge  shown  at  lOA  will 
not  be  satisfactory  for  the  measurement  of  ca- 
pacitances smaller  than  about  1000  gg.fd.  For 
the  measurement  of  capacitances  from  a few 
micro-microfarads  to  about  0.001  /rfd.  a Wag- 
ner grounded  substitution  capacitance  bridge  of 
the  type  shown  in  figure  lOB  will  be  found 
satisfactory.  The  ratio  arms  Ra  and  Rb  should 
be  of  the  same  value  within  1 per  cent;  any 
value  between  2500  and  10,000  ohms  for 
both  will  be  satisfactory.  The  two  resistors 
Ro  and  Rd  should  be  1000-ohm  wire-wound 
potentiometers.  Cs  should  be  a straight-line 
capacitance  capacitor  with  an  accurate  vernier 
dial;  500  to  1000  g/rfd.  will  be  satisfactory. 
Cv  can  be  a two  or  three  gang  broadcast  ca- 
pacitor from  700  to  1000  ggfd.  maximum  ca- 
pacitance. 

The  procedure  for  making  a measurement 
is  as  follows;  The  unknown  capacitor  Cx  is 
placed  in  parallel  with  the  standard  capacitor 
Cs.  The  ]Vag»er  ground  Rn  is  varied  back  and 
forth  a small  amount  from  the  center  of  its 
range  until  no  signal  is  heard  in  the  phones 
with  the  switch  S in  the  center  position.  Then 
the  switch  S is  placed  in  either  of  the  two  out- 
side positions,  Co  is  adjusted  to  a capaci- 
tance somewhat  greater  than  the  assumed  value 
of  the  unknown  Cx,  and  the  bridge  is  brought 
into  balance  by  variation  of  the  standard  ca- 
pacitor Cs.  It  may  be  necessary  to  cut  some 
resistance  in  at  Ro  and  to  switch  to  the  other 
outside  position  of  S before  an  exaa  balance 
can  be  obtained.  The  setting  of  Cs  is  then 
noted,  Cx  is  removed  from  the  circuit  ( but  the 
leads  which  went  to  it  are  not  changed  in  any 
way  which  would  alter  their  mumal  capaci- 
tance), and  Cs  is  readjusted  until  balance  is 
again  obtained.  The  difference  in  the  two  set- 
tings of  Cs  is  equal  to  the  capacitance  of  the 
unknown  capacitor  Cx. 

There  are  many  other  types  of  a-c  bridge 
circuits  in  common  use  for  measuring  indua- 
ance  with  a capacitance  standard,  frequency 


Zx- IMPEDANCE  BEING  MEASURED,  Rs  = RESISTANCE  COMPONENT  OF  Zs 
Zs=  IMPEDANCE  OF  STANDARD,  Xx^REACTANCE  COMPONENT  OF  Zx 
Rx=RESISTANCE  COMPONENT  OF  Zx,  Xs'REACTAnCE  COMPONENT  OF  Zs 


TWO  A-C  BRIDGE  CIRCUITS 


The  operaiion  of  these  bridges  is  esseniiafly 
the  some  as  those  of  figure  9 except  that 
a.e.  is  fed  into  the  bridge  instead  of  d.e.  and 
a pair  of  phones  is  used  os  the  indicator  in- 
stead of  the  gaiYonometer.  The  bridge  shown 
at  (A)  can  be  used  for  the  measurement  of 
resistance,  but  it  is  usually  used  for  the 
measurement  of  the  impedance  and  reactance 
of  coils  and  capacitors  at  frequencies  from 
200  to  1000  cycles.  The  bridge  shown  at  (B) 
is  used  for  the  measurement  of  small  values 
of  capacitance  by  the  substitution  method. 
Full  description  of  the  operation  of  both 
bridges  is  given  in  the  accompanying  text. 


in  terms  of  resistance  and  capacitance,  and 
so  forth.  Terman’s  Radio  Engineers’  Handbook 
gives  an  excellent  discussion  of  common  types 
of  a-c  bridge  circuits. 

34-4  Frequency 
Measurements 

All  frequency  measurement  within  the 
United  States  is  based  on  the  transmissions  of 
Station  WWV  of  the  National  Bureau  of 
Standards.  This  station  operates  continuously 
on  frequencies  of  2.5,  5,  10,  15,  20,  25,  30, 
and  35  Me.  The  carriers  of  those  frequencies 
below  30  Me.  are  modulated  alternately  by 
a 440-cycle  tone  or  a 600-cycle  tone  for  pe- 
riods of  four  minutes  each.  This  tone  is  in- 
terrupted at  the  beginning  of  the  59th  minute 


730-B  Test  Equipment 


THE  RADIO 


Cl — 100-^L/j.fd.  air  trimmer 
CfyCj — 0.0003-jifd.  midget  mica 
Ci — 50-fi/j.fd.  midget  mica 
Cs — 0.002-mW.  midget  mica 
Ri,  Rs — 100,000  ohms  1/2  '♦'Olf 
Li — 10-mh.  shielded  r-f  choke 
Ls — 2.1 -mh.  r-f  choke 
X — 100-kc,  crystal 


SCHEMATIC  OF  A 100-KC.  FREQUENCY  SPOTTER 


of  each  hour  and  each  five  minutes  thereafter 
for  a period  of  precisely  one  minute.  Green- 
wich Civil  Time  is  given  in  code  during  these 
one-minute  intervals,  followed  by  a voice  an- 
nouncement giving  Eastern  Standard  Time. 
The  accuracy  of  all  radio  and  audio  frequen- 
cies is  better  than  one  part  in  50,000,000. 
A 5000  microsecond  pulse  ( 5 cycles  of  a 1000- 
cycle  wave)  may  be  heard  as  a tick  for  every 
second  except  the  59th  second  of  each  minute. 

These  standard-frequency  transmissions  of 
station  WWV  may  be  used  for  accurately  de- 
termining the  limits  of  the  various  amateur 
bands  with  the  aid  of  the  station  communica- 
tions receiver  and  a 50-kc.,  100-kc.,  or  200- 
kc.  band-edge  spotter.  The  low  frequency  oscil- 
lator may  be  self-excited  if  desired,  but  low 
frequency  standard  crystals  have  become  so 
relatively  inexpensive  that  a reference  crystal 
may  be  purchased  for  very  little  more  than 
the  cost  of  the  components  for  a self-excited 
oscillator.  The  crystal  has  the  additional  advan- 
tage that  it  may  be  once  set  so  that  its  har- 
monics are  at  zero  beat  with  WWV  and  then 
left  with  only  an  occasional  check  to  see  that 
the  frequency  has  not  drifted  more  than  a 
few  cycles.  The  self-excited  oscillator,  on  the 
other  hand,  must  be  monitored  very  frequent- 
ly to  insure  that  it  is  on  frequency. 

Using  a To  use  a frequency  spotter  it  is 
Frequency  only  necessary  to  couple  the  out- 
Spotl-er  put  of  the  oscillator  unit  to  the 
antenna  terminal  of  the  receiver 
through  a very  small  capacitance  such  as  might 
be  made  by  twisting  two  pieces  of  insulated 
hookup  wire  together.  Station  WWV  is  then 
tuned  in  on  one  of  its  harmonics,  15  Me. 
will  usually  be  best  in  the  daytime  and  5 or 
10  Me.  at  night,  and  the  trimmer  adjustment 
on  the  oscillator  is  varied  until  zero  beat  is 
obtained  between  the  harmonic  of  the  oscil- 
lator and  WWV.  With  a crystal  reference 
oscillator  no  difficulty  will  be  had  with  using 


the  wrong  harmonic  of  the  oscillator  to  ob- 
tain the  beat,  but  with  a self-excited  oscillator 
it  will  be  wise  to  insure  that  the  reference 
oscillator  is  operating  exactly  on  50,  100,  or 
200  kc.  (whichever  frequency  has  been 
chosen)  by  making  sure  that  zero  beat  is  ob- 
tained simultaneously  on  all  the  frequencies 
of  WWV  that  can  be  heard,  and  by  noting 
whether  or  not  the  harmonics  of  the  oscillator 
in  the  amateur  bands  fall  on  the  approximate 
calibration  marks  of  the  receiver. 

A simple  frequency  spotter  is  diagrammed 
in  figure  11. 

34-5  Anfenna  and 
Transmission 
Line  Measurements 

The  degree  of  adjustment  of  any  amateur 
antenna  can  be  judged  by  the  study  of  the 
standing- wave  ratio  on  the  transmission  line 
feeding  the  antenna.  Various  types  oi^  i nstru- 
ments  have  been  designed  to  measure  the 
s.r.w.  present  on  the  transmission  line,  or  to 
measure  the  actual  radiation  resistance  of  the 
antenna  in  question.  The  most  important  of 
these  instruments  are  the  slotted  line,  the 
bridge-type  s-w-r  meter,  and  the  antennascope. 

The  Slotted  Line  It  is  obviously  impractical 
to  measure  the  voltage- 
standing-wave ratio  in  a length  of  coaxial  line 
since  the  voltages  and  currents  inside  the  line 
are  completely  shielded  by  the  outer  conductor 
of  the  cable.  Hence  it  is  necessary  to  insert 
some  type  of  instrument  into  a section  of  the 
line  in  order  to  be  able  to  ascertain  the  con- 
ditions which  are  taking  place  inside  the 
shielded  line.  Where  measurements  of  a high 
degree  of  accuracy  are  required,  the  slotted 
line  is  the  instrument  most  frequently  used. 
Such  an  instrument,  diagrammed  in  figure  12, 
is  an  item  of  test  equipment  which  could  be 
constructed  in  a home  workshop  which  in- 
cluded a lathe  and  other  metal  working  tools. 


HANDBOOK 


Antenna  Measurements  731 


Figure  12 

DIAGRAMMATIC  REPRESENTATION 
OF  A SLOTTED  LINE 

The  conductor  ratios  in  the  slotted  fine,  fn> 
eluding  the  tapered  end  sections  should  be 
such  that  the  characteristic  impedance  of  the 
equipment  is  the  same  as  that  of  the  trans- 
mission line  with  which  the  equipment  is  to 
be  used.  The  indicating  instrument  may  be 
operated  by  the  d-c  output  of  the  rectifier 
coupled  to  the  probe,  or  it  may  be  operated 
by  the  a-c  components  of  the  rectified  signal 
if  the  signal  generator  or  transmitter  is  am- 
plitude modulated  by  a constant  percentage. 


Commercially  built  slotted  lines  are  very  ex- 
pensive since  they  are  constructed  with  a high 
degree  of  accuracy  for  precise  laboratory  work. 
The  slotted  line  consists  essentially  of  a 
section  of  air-dielectric  line  having  the  same 
characteristic  impedance  as  the  transmission 
line  into  which  it  is  inserted.  Tapered  fittings 
for  the  transmission  line  connectors  at  each 
end  of  the  slotted  line  usually  ate  requited  due 
to  differences  in  the  diameters  of  the  slotted 
line  and  the  line  into  which  it  is  inserted.  A 
narrow  slot  from  Vs-inch  to  14-inch  in  width 
is  cut  into  the  outer  conductor  of  the  line.  A 
probe  then  is  insetted  into  the  slot  so  that  it 
is  coupled  to  the  field  inside  the  line.  Some 
sort  of  accurately  machined  track  or  lead  screw 
must  be  provided  to  insure  that  the  probe 
maintains  a constant  spacing  from  the  inner 
conductor  as  it  is  moved  from  one  end  of  the 
slotted  line  to  the  other.  The  probe  usually 
includes  some  type  of  rectifying  element 
whose  output  is  fed  to  an  indicating  instru- 
ment alongside  the  slotted  line. 

The  unfortunate  part  of  the  slocted-line  sys- 
tem of  measurement  is  that  the  line  must  be 
somewhat  over  one-half  wavelength  long  at 
the  test  frequency,  and  for  best  results  should 
be  a full  wavelength  long.  This  requirement  is 
easily  met  at  frequencies  of  420  Me.  and  above 
where  a full  wavelength  is  28  inches  or  less. 
But  for  the  lower  frequencies  such  an  instru- 
ment is  mechanically  impracticable. 

Bridge-Type  The  bridge  type  of  standing- 

Standing-Wave  wave  indicator  is  used  quite 
Indicators  generally  for  making  meas- 

urements on  commercial  co- 


RESISTOR-BRIDGE 
STANDING-WAVE  INDICATOR 


This  type  of  test  equipment  is  suitable 
for  use  with  coaxial  feed  lines. 


C, — O.OOJ-Mfd.  midget  ceramic  capacitor 
Cf  Cj — .007  disc  ceramic 
Ri,  Ri — 22-ohm  2-wati  carbon  resistors 
Rs — Resistor  equal  in  resistance  to  the  charac- 
teristic impedance  of  the  coaxial  transmis- 
sion line  to  be  used  (J  watt) 

R/. — 5000-ohm  wire-wound  potentiometer 
Rs — 70,000-ohm  1-watt  resistor 
RFC — R-f  choke  suitable  for  operation  at  the 
measurement  frequency 


axial  transmission  lines.  A simplified  version 
is  available  from  M.  C.  Jones  Electronics  Co., 
Bristol,  Conn.  ("Micro-Match”). 

One  type  of  bridge  standing-wave  indicator 
is  diagrammed  in  figure  13.  This  type  of  in- 
strument compares  the  electrical  impedance  of 
the  transmission  line  with  that  of  the  resistor 
Rs  which  is  included  within  the  unit.  Experi- 
ence with  such  units  has  shown  that  the  re- 
sistor Rs  should  be  a good  grade  of  non-induc- 
tive carbon  type.  The  Ohmite  "Little  Devil" 
type  resistor  in  the  2 -watt  rating  has  given 
good  performance.  The  resistance  at  Rs  should 
be  equal  to  the  characteristic  impedance  of 
the  antenna  transmission  line.  In  other  words, 
this  resistor  should  have  a value  of  52  ohms 
for  lines  having  this  characteristic  impedance 
such  as  RG-8/U  and  RG-58/U.  For  use  with 
lines  having  a nominal  characteristic  im- 
pedance of  70  ohms,  a selected  "68  ohm”  re- 
sistor having  an  actual  resistance  of  70  ohms 
may  be  used. 

Balance  within  the  equipment  is  checked  by 
mounting  a resistor,  equal  in  value  to  the 
nominal  characteristic  impedance  of  the  line 
to  be  used,  on  a coaxial  plug  of  the  type  used 
on  the  end  of  the  antenna  feed  line.  Then 
this  plug  is  inserted  into  the  input  receptacle 
of  the  instrument  and  a power  of  2 to  4 watts 
applied  to  the  output  receptacle  on  the  desired 
frequency  of  operation.  Note  that  the  signal 
is  passed  through  the  bridge  in  the  direction 


732  Test  Equipment 


THE  RADIO 


READING  ON  0*1  INSTRUMENT,  OR  FACTOR  TIMES  FULL  SCALE 
(MAGNITUDE  OF  REFLECTION  COEFFICIENT,  A ) 

Figure  14 

RELATION  BETWEEN  STANDING- 
WAVE  RATIO  AND  REFLECTION 
COEFFICIENT 

This  chart  may  he  used  to  convert  reflection^ 
coefficient  indications  such  as  are  obtained 
with  a bridge-type  standing-wave  indicator 
or  an  indicating  twin  lamp  into  values  of 
standing-wave  ratio. 


REFLECTOMETER 

OUTER  SHELL 


COMPONENT  PARTS 

END  DISC  = Z 3/6"  DIAMETER  X 1/4"  (Z  ffeQ.) 

OUTER  SHELLS  2 3/8"  (.0.  X 6"  (/ 

ALIGNMENT  ROD*  1/4"  DIAMETER  X S !/2"  (2  REQ-  ) 

INNER  CONDUCTOR  = 1/2 " Dl  AM ETER  X 5 1/8",  TA  PER  ENOS 
TO  SOLDER  TO  RECEPTACLES  {Z  REQ.) 

RECEPTACLES*  SO-239  {Z  REQ.)  (Jl,J2) 

BINDING  POSTS*  {3  REQ.) 

Figure  15 

SCHEMATIC,  REFLECTOMETER 

Di,  Di — Crystal  diode,  1N34A  or  1N82 
Ri—270  ohm,  1 wait  composition  resistor. 

IRC  type  BTA,  matched  pair. 

M — 0-1  d.c.  milliammeter 

Jg — Coaxial  receptacle,  50-239. 


opposite  to  normal  for  this  test.  The  resistor 
Ri  is  adjusted  for  full-scale  deflection  on  the 
0-100  microammeter.  Then  the  plugs  are  re- 
versed so  that  the  test  signal  passes  through 
the  instrument  in  the  direction  indicated  by 
the  arrow  on  figure  13,  and  the  power  level 
is  maintained  the  same  as  before.  If  the  test 
resistor  is  matched  to  Ra,  and  stray  capacitances 
have  been  held  to  low  values,  the  indication 
on  the  milliammeter  will  be  very  small.  The 
test  plug  with  its  resistor  is  removed  and  the 
plug  for  the  antenna  transmission  line  is  in- 
serted. The  meter  indication  now  will  read 
the  reflection  coefficient  which  exists  on  the 
antenna  transmission  line  at  the  point  where 
the  indicator  has  been  inserted.  From  this 
reading  of  reflection  coefficient  the  actual 
standing-wave  ratio  on  the  transmission  line 
may  be  determined  by  reference  to  the  chart 
of  figure  14. 

Measuremenrs  of  this  type  are  quite  helpful 
in  determining  whether  or  not  the  antenna  is 
presenting  a good  impedance  match  to  the 
transmission  line  being  used  to  feed  it.  How- 
ever, a test  instrument  of  the  type  shown  in 


figure  13  must  be  inserted  into  the  line  for  a 
measurement,  and  then  removed  from  the  line 
when  the  equipment  is  to  be  operated.  Also, 
the  power  input  to  the  line  feeding  the  input 
terminal  of  the  standing-wave  indicator  must 
not  exceed  4 watts.  The  power  level  which  the 
unit  can  accept  is  determined  by  the  dissipa- 
tion limitation  of  resistors  Ri  plus  Rs. 

It  is  also  important,  for  satisfactory  opera- 
tion of  the  test  unit,  that  resistors  Ri  and  Rj 
be  exactly  equal  in  value.  The  actual  resist- 
ance of  these  two  is  not  critically  important, 
and  deviations  up  to  10  per  cent  from  the 
value  given  in  figure  13  will  be  satisfactory. 
But  the  two  resistors  must  have  the  same  value, 
whether  they  are  both  21  ohms  or  24  ohms, 
or  some  value  in  between. 

34-6  A Simple 

Coaxial  Reflecfometer 

The  reflectometer  is  a short  section  of  co- 
axial transmission  line  containing  two  r-f  volt- 
meters. One  voltmeter  reads  the  incident  com- 
ponent of  voltage  in  the  line,  and  the  other 


HANDBOOK 


Reflectometer  733 


Figure  16 

INTERIOR  VIEW  OF  COAXIAL  REFLECTOMETER 

The  Reflectometer  is  a short  section  of  transmission  line  containing  two  r-f  voltmeters.  Center  conductor 
of  line  is  a section  of  brass  rod  soldered  to  center  pins  of  input  and  output  receptacles.  At  either  end 
of  unit  are  the  crystal  diodes,  bypass  capacitors  and  terminals.  Diode  load  resistors  are  at  center  of 
instrument,  grounded  to  brass  alignment  rod. 


reads  the  reflected  component.  The  magnitude 
of  standing  wave  ratio  on  the  transmission 
line  is  the  ratio  of  the  incident  component 
to  the  reflected  component,  as  shown  in  fig- 
ure 14.  In  actual  use,  calibration  of  the  re- 
flectometer is  not  required  since  the  relative 
reading  of  reflected  power  indicates  the  de- 
gree of  match  or  mis-match  and  all  antenna 
and  transmission  line  adjustments  should  be 
conducted  so  as  to  make  this  reading  as  low 
as  possible,  regardless  of  its  absolute  value. 

The  actual  meter  readings  obtained  from 
the  device  are  a function  of  the  operating  fre- 
quency, the  sensitivity  of  the  instrument  being 
a function  of  transmitter  power,  increasing 
rapidly  as  the  frequency  of  operation  is  in- 
creased. However,  the  reflectometer  is  inval- 
uable in  thar  it  may  be  left  permanently  in 
the  transmission  line,  regardless  of  the  power 
output  level  of  the  transmitter.  It  will  indi- 
cate the  degree  of  reflected  power  in  the  an- 
tenna system,  and  at  the  same  time  provide 
a visual  indication  of  the  power  output  of 
the  transmitter. 

Reflectometer  The  circuit  and  assembly  in- 
Circuit  formation  for  the  reflectometer 

are  given  in  figure  15.  Two 
diode  voltmeters  are  coupled  back-to-back  to 
a short  length  of  transmission  line.  The  com- 
bined inductive  and  capacative  pickup  between 
each  voltmeter  and  the  line  is  such  that  the 
incident  component  of  the  line  voltage  is  bal- 
anced out  in  one  case  and  the  induaive  com- 
ponent is  balanced  out  in  the  other  case.  Each 
voltmeter,  therefore,  reads  only  one  wave-com- 


ponent. Careful  attention  to  physical  symmetry 
of  the  assembly  insures  accurate  and  complete 
separation  of  the  voltage  components  by  the 
two  voltmeters.  The  outputs  of  the  two  volt- 
meters may  be  selected  and  read  on  an  ex- 
ternal meter  connected  to  the  terminal  posts 
of  the  reflectometer. 

Each  r-f  voltmeter  is  composed  of  a load 
resistor  and  a pickup  loop.  The  pickup  loop 
is  positioned  parallel  to  a section  of  transmis- 
sion line  permitting  both  inductive  and  ca- 
pacative coupling  to  exist  between  the  center 
conductor  of  the  line  and  the  loop.  The  di- 
mensions of  the  center  conduaor  and  the 
outer  shield  of  the  reflectometer  are  chosen 
so  that  the  instrument  impedance  closely 
matches  that  of  the  transmission  line. 

Reflectometer  A view  of  the  interior  of  the 
Construction  reflectometer  is  shown  in  fig- 

ure 16.  The  coaxial  input  and 
output  connectors  of  the  instrument  ate 
mounted  on  machined  brass  discs  that  are 
held  in  place  by  brass  alignment  rods,  tapped 
at  each  end.  The  center  conductor  is  machined 
from  a short  section  of  brass  rod,  tapered  and 
drilled  at  each  end  to  fit  over  the  center  pin 
of  each  coaxial  receptacle.  The  end  discs,  the 
rods,  and  the  center  conductor  should  be  sil- 
ver plated  before  assembly.  When  the  center 
conductor  is  placed  in  position,  it  is  soldered 
at  each  end  to  the  center  pin  of  the  coaxial 
receptacles. 

One  of  the  alignment  rods  is  drilled  and 
tapped  for  a 6-32  bolt  at  the  mid-point,  and 
the  end  discs  are  drilled  to  hold  Vi -inch 


734  Test  Equipment 


THE  RADIO 


ceramic  insulators  and  binding  posts,  as  shown 
in  the  photograph.  The  load  resistors,  crystal 
diodes,  and  bypass  capacitors  are  finally 
mounted  in  the  assembly  as  the  last  step. 

The  two  load  resistors  should  be  measured 
on  an  ohmmeter  to  ensure  that  the  resistance 
values  are  equal.  The  exact  value  of  resistance 
is  unimportant  as  long  as  the  two  resistors 
are  equal.  The  diodes  should  also  be  checked 
on  an  ohmmeter  to  make  sure  that  the  front 
resistances  and  back  resistances  are  balanced 
between  the  units.  Care  should  be  taken  dur- 
ing soldering  to  ensure  the  diodes  and  resis- 
tors are  not  overheated.  Observe  that  the  re- 
sistor leads  are  of  equal  length  and  that  each 
half  of  the  assembly  is  a mirror-image  of  the 
other  half.  The  body  of  the  resistor  is  spaced 
about  V^-inch  away  from  the  center  conduc- 
tor. 

Testing  the  The  instrument  can  be  ad- 
Reflectometer  justed  on  the  28  Me.  band. 

An  r-f  source  of  a few  watts 
and  nonreactive  load  are  required.  The  con- 
struction of  the  reflectometer  is  such  that  it 
will  work  well  with  either  52-  or  72-ohm 
coaxial  transmission  lines.  A suitable  dummy 
load  for  the  52-ohm  line  can  be  made  of  four 
220  ohm,  2 watt  composition  resistors  {Ohm- 
ite  "Little  Devil")  connected  in  parallel.  Clip 
the  ieads  of  the  resistors  short  and  mount 
them  on  a coaxial  plug.  This  assembly  pro- 
vides an  eight  watt,  55  ohm  load,  suitable 
for  use  at  30  Me.  If  an  accurate  ohmmeter 
is  at  hand,  the  resistors  may  be  hand  picked 
to  obtain  four  208  ohm  units,  thus  making 
the  dummy  load  resistors  exactly  52  ohms. 
For  all  practical  purposes,  the  55  ohm  load 
is  satisfactory.  A 75  ohm,  eight  watt  load 
resistor  may  be  made  of  four  300  ohm,  2 
watt  composition  resistors  connected  in  paral- 
lel. 

R-f  power  is  coupled  to  the  reflectometer 
and  the  dummy  load  is  placed  in  the  "output” 
receptacle.  The  indicator  meter  is  switched  to 
the  "reflected  power”  position.  The  meter  read- 
ing should  be  almost  zero.  It  may  be  brought 
to  zero  by  removing  the  case  of  the  instru- 
ment and  adjusting  the  position  of  the  load 
resistor.  The  actual  length  of  wire  in  the  resis- 
tor lead  and  its  positioning  determine  the  meter 
null.  Replace  the  case  before  power  is  applied 
to  the  reflectometer.  The  reflectometer  is  now 
reversed  and  power  is  applied  to  the  "output” 
receptacle,  with  a dummy  load  attached  to 
the  "input”  receptacle.  The  second  voltmeter 
(forward  power)  is  adjusted  for  a null  read- 
ing of  the  meter  in  the  same  manner. 


If  a reflected  reading  of  zero  is  not  obtain- 
able, the  harmonic  content  of  the  r-f  source 
might  be  causing  a slight  residual  meter  read- 
ing. Coupling  the  reflectometer  to  the  r-f 
source  through  a tuned  circuit  ("antenna 
tuner”)  will  remove  the  offending  harmonic 
and  permit  an  accurate  null  indication.  Be 
sure  to  hold  the  r-f  input  power  to  a low 
value  to  prevent  overheating  the  dummy  load 
resistors. 

Using  the  The  bridge  may  be  used  up  to 
Reflectometer  150  Me.  It  is  placed  in  the 
transmission  line  at  a conveni- 
ent point,  preferably  before  any  tuner,  balun, 
or  TVI  filter.  The  indicator  should  be  set  to 
read  forward  power,  with  a maximum  of  re- 
sistance in  the  circuit.  Power  is  applied  and 
the  indicator  resistor  is  adjusted  for  a full 
scale  reading.  The  switch  is  then  thrown  to 
read  reflected  power  (indicated  as  A,  figure 
14).  Assume  that  the  forward  power  meter 
reading  is  1.0  and  the  reflected  power  read- 
ing is  0.5.  Substituting  these  values  in  the 
SWR  formula  of  figure  14  shows  the  SWR 
to  be  3.  If  forward  power  is  always  set  to 
1.0  on  the  meter,  the  reflected  power  (A) 
can  be  read  directly  from  the  curve  of  figure 
14  with  little  error. 

If  the  meter  is  adjusted  so  as  to  provide  a 
half-scale  reading  of  the  forward  power,  the 
reflectometer  may  be  used  as  a transmitter 
power  output  meter.  Tuning  adjustments  may 
then  be  undertaken  to  provide  greatest  meter 
reading. 

34-7  Measurement's 
on  Balanced 
Transmission  Lines 

Measurements  made  on  balanced  transmis- 
sion lines  may  be  conducted  in  the  same  man- 
ner as  those  made  on  coaxial  lines.  In  the 
case  of  the  coaxial  lines,  care  must  be  taken 
to  prevent  flow  of  r-f  current  on  the  outer 
surface  of  the  line  as  this  unwanted  compo- 
nent will  introduce  errors  in  measurements 
made  on  the  line.  In  like  fashion,  the  cur- 
rents in  a balanced  transmission  line  must 
be  180  degrees  out  of  phase  and  balanced 
with  respect  to  ground  in  order  to  obtain  a 
realistic  relationship  between  incident  and  re- 
flected power.  This  situation  is  not  always 
easy  to  obtain  in  practice  because  of  the  prox- 
imity effects  of  metallic  objects  or  the  earth 
to  the  transmission  line.  All  transmission  line 
measurements,  therefore,  should  be  conducted 
with  the  realization  of  the  physical  limitations 


HANDBOOK 


S.W.R.  Indicator  735 


SKETCH  OF  THE  "TWIN-LAMP" 
TYPE  OF  S-W-R  INDICATOR 


The  short  section  oi  line  with  lamps  at  each 
end  usually  is  taped  to  the  main  transmission 
line  with  plastic  electrical  tape. 


of  the  equipment  and  the  measuring  technique 
that  is  being  used. 

Measurements  on  One  of  the  most  satisfaaory 
Molded  Parallel-  and  least  expensive  devices 
Wire  Lines  for  obtaining  a rough  idea 

of  the  standing-wave  ratio 
on  a transmission  line  of  the  molded  parallel- 
wire  type  is  the  twin-lamp.  This  ingenious  in- 
strument may  be  constructed  of  new  compon- 
ents for  a total  cost  of  about  25  cents;  this 
fact  alone  places  the  twin-lamp  in  a class  by 
itself  as  far  as  test  instruments  ate  concerned. 

Figure  17  shows  a sketch  of  a twin-lamp  in- 
dicator. The  indicating  portion  of  the  system 
consists  merely  of  a length  of  300-ohm  Twin- 
Lead  about  10  inches  long  with  a dial  lamp  at 
each  end.  In  the  unit  illustrated  the  dial  lamps 
ate  standard  6.3-volt  150-ma.  bayonet-base 
lamps.  The  lamps  are  soldered  to  the  two  leads 
at  each  end  of  the  short  section  of  Twin-Lead. 

To  make  a measurement  the  short  section 
of  line  with  the  lamps  at  each  end  is  merely 
taped  to  the  section  of  Twin-Lead  (or  other 
similar  transmission  line)  running  from  the 
transmitter  or  from  the  antenna  changeover  re- 
lay to  the  antenna  system.  When  there  are  no 
standing  waves  on  the  antenna  transmission 
line  the  lamp  toward  the  transmitter  will  light 
while  the  one  toward  the  antenna  will  not 
light.  With  300-ohm  Twin-Lead  running  from 
the  antenna  changeover  relay  to  the  antenna, 
and  with  about  200  watts  input  on  the  28-Mc. 
band,  the  dial  lamp  toward  the  transmitter  will 
light  nearly  to  full  brilliancy.  With  a standing- 
wave  ratio  of  about  1.5  to  1 on  the  transmis- 
sion line  to  the  antenna  the  lamp  toward  the 
antenna  will  just  begin  to  light.  With  a high 
standing-wave  ratio  on  the  antenna  feed  line 
both  lamps  will  light  nearly  to  full  brilliancy. 
Hence  the  instrument  gives  an  indication  of 
relatively  low  standing  waves,  but  when  the 


Figure  18 

OPERATION  OF  THE  "TWIN  LAMP" 
INDICATOR 

Showing  current  flow  resulting  from  inductive 
and  capacitive  fields  in  a "twin  lamp"  attached 
to  o line  with  a low  standing-wave  ratio. 


standing-wave  ratio  is  high  the  twin-lamp 
merely  indicates  that  they  are  high  without 
giving  any  idea  of  the  actual  magnitude. 

Operation  of  The  twin  lamp  operates  by 
the  Twin-Lamp  virtue  of  the  fact  that  the  ca- 
pacitive and  inductive  cou- 
pling of  the  wire  making  up  one  side  of  the 
twin-lamp  is  much  greater  to  the  transmission- 
line lead  immediately  adjacent  to  it  than  to 
the  transmission-line  lead  on  the  other  side. 
The  same  is  of  course  true  of  the  wire  on  the 
other  side  of  the  twin-lamp  and  the  transmis- 
sion-line lead  adjacent  to  it.  A further  condi- 
tion which  must  be  met  for  the  twin-lamp  to 
operate  is  that  the  section  of  line  making  up 
the  twin-lamp  must  be  short  with  respect  to  a 
quarter  wavelength.  Then  the  current  due  to 
capacitive  coupling  passes  through  both  lamps 
in  the  same  direction,  while  the  current  due  to 
inductive  coupling  between  the  leads  of  the 
twin-lamp  and  the  leads  of  the  antenna  trans- 
mission line  passes  through  the  two  lamps  in 
opiX)site  directions.  Hence,  in  a line  without 
reflections,  the  two  currents  will  cancel  in 
one  lamp  while  the  other  lamp  is  lighted  due 
to  the  sum  of  the  currents  (figute  18). 

The  basic  fact  which  makes  the  twin-lamp  a 
directional  coupler  is  a result  of  the  condition 
whereby  the  capacitive  coupling  is  a scalar 
aaion  not  dependent  upon  the  direction  of  the 
waves  passing  down  the  line,  yet  the  inductive 
coupling  is  a vector  action  which  is  dependent 
upon  the  direction  of  wave  propagation  down 
the  line.  Thus  the  capacitive  current  is  the 
same  and  is  in  the  same  phase  for  energy 
travelling  in  either  direction  down  the  line. 
But  the  inductive  current  travels  in  one  direc- 
tion for  energy  travelling  in  one  direction  and 
in  the  other  direaion  for  energy  going  the 
other  direction.  Hence  the  two  currents  add  at 
one  end  of  the  line  for  a wave  passing  toward 
the  antenna,  while  the  currents  add  at  the 
other  end  of  the  twin-lamp  for  the  waves  re- 
flected from  the  antenna.  When  the  waves  are 


736  Test  Equipment 


strongly  reflected  upon  reaching  the  antenna, 
the  reflected  wave  is  nearly  the  same  as  the 
direct  wave,  and  both  lamps  will  light.  This 
condition  of  strong  reflection  from  the  antenna 
system  is  that  which  results  in  a high  stand- 
ing-wave ratio  on  the  antenna  feed  line. 

Use  of  the  The  twin-lamp  is  best 

Twin  Lamp  with  suited  for  use  with  an- 

Vorious  Feed  Lines  tenna  transmission  lines 
of  the  flat  ribbon  type. 
Lines  with  a high  power  rating  are  available 
in  impedances  of  75  ohms,  205  ohms,  and  300 
ohms  in  the  flat  ribbon  type.  In  addition,  one 
manufacturer  makes  a 300-ohm  line  in  two 
power-level  ratings  with  a tubular  cross  sec- 
tion. A twin-lamp  made  from  flat  300-ohm 
line  may  be  used  with  this  tubular  line  by 
taping  the  twin-lamp  tightly  to  the  tubular 
line  so  that  the  conduaors  of  the  twin-lamp 
are  as  close  as  possible  on  each  side  of  the  con- 
ductors of  the  antenna  line. 

34-8  A "Balanced" 

SWR  Bridge 

Two  resistor-type  standing  wave  indicators 
may  be  placed  "back-to-back”  to  form  a SWR 
bridge  capable  of  being  used  on  two  wire 
balanced  transmission  lines.  Such  a bridge  is 
shown  in  figures  19  and  21.  The  schematic  of 
such  an  instrument  (figure  20)  may  be  com- 
pared to  two  of  the  simple  bridges  shown  in 
figure  13.  When  the  dual  bridge  circuit  is 
balanced  the  meter  reading  is  zero.  This  state 
is  reached  when  the  line  currents  are  equal 
and  exactly  180  degrees  out  of  phase  and  the 
SWR  is  unity. 

As  the  condition  of  the  line  departs  from 
the  optimum,  the  meter  of  the  bridge  will 


Figure  19 

SWR  BRIDGE  FOR  BALANCED 
TRANSMISSION  LINE 

A double  bridge  can  be  used  for  two  wire 
transmission  lines.  Bridge  is  inserted  in  line 
and  may  be  driven  with  grid-dip  oscillator  or 
other  low  power  r-f  source. 


show  the  degree  of  departure.  When  the  line 
currents  are  balanced  and  180  degrees  out  of 
phase,  the  meter  will  read  the  true  value  of 
standing  wave  ratio  on  the  line.  If  these  con- 
ditions are  not  met,  the  reading  is  not  abso- 
lute, merely  giving  an  indication  of  the  degree 
of  mis-match  in  the  line.  This  handicap  is  not 
important,  since  the  relative,  not  the  absolute, 
degree  of  mis-match  is  sufficient  for  transmis- 
sion line  adjustments  to  be  made. 

Bridge  A suggested  method  of  con- 

ConstrucKon  struction  of  the  balanced  bridge 
is  illustrated  in  figures  19  and 
21.  The  unit  is  constructed  within  a box  mea- 
suring 4"  X 6"  X 2”  in  size.  The  0 - 200  d.c. 
microammeter  is  placed  in  the  center  of  the 
4"  X 6"  side  of  the  case.  The  input  and  output 
connectors  of  the  instrument  are  placed  on 
each  end  of  the  box  and  the  internal  wiring 
is  arranged  so  that  the  transmission  line,  in 
effect,  passes  in  one  side  of  the  box  and  out 
the  other  with  as  little  discontinuity  as  possi- 
ble. The  input  and  output  terminals  are  mount- 
ed on  phenolic  plates  placed  over  large  cut- 
outs in  the  ends  of  the  box,  thus  reducing 
circuit  capacity  to  ground  to  a minimum  value. 
The  "SWR-CAL”  switch  Si  is  located  on  one 
side  of  the  meter  and  the  "Calibrate”  poten- 
tiometer Ri  is  placed  on  the  opposite  side. 

The  transmission  line  within  the  unit  is 


Rl 


Figure  20 

schematic  of  bridge  for 

BALANCED  LINES 

M — 0-200  d-c  microammeter 
Rl — Note:  Six  250  ohm  resistors  are  composi- 
tion, non-inductive  units.  IRC  type  BT,  or 
Ohmite  "Little  Devil"  1-watt  resistors  may 
be  used,  (see  text) 

Si — DPDT  rotary  switch.  Centralab  type  1464 


Figure  21 

INTERIOR  VIEW  OF 
BALANCED  BRIDGE 
SHOWING  PARTS 
PLACEMENT 

Diode  rectifiers  are  placed 
at  right  angles  to  the 
short  section  of  trans- 
mission  line.  Both  sides 
of  bridge  are  balanced 
to  ground  by  virtue  of 
symmetrica/  construction 
of  unit. 


broken  by  two  250  ohm  composition  resistors 
and  switch  Si.  The  line  segments  are  made  of 
short  pieces  of  #10  copper  wire,  running  be- 
tween the  various  components.  Spacing  be- 
tween the  wires  is  held  close  to  three  inches 
to  approximate  a 500-600  ohm  line. 

The  small  components  of  the  bridge  ate 
placed  symmetrically  about  the  250  ohm  series 
line  resistors  and  the  calibrating  potentiometer, 
as  can  be  seen  in  figure  21.  Exact  parts  place- 
ment is  not  critical,  except  that  the  crystal 
diodes  should  be  placed  at  right  angles  to 
the  wires  of  the  transmission  line  to  reduce 
capacative  pickup.  The  two  r-f  chokes  should 
then  be  placed  at  right  angles  to  the  diodes.  • 

The  six  250  ohm  resistors  are  checked  on  an 
ohmmeter  and  should  be  hand-picked  to  ob- 
tain units  that  ate  reasonably  close  in  value. 
If  it  is  desired  to  use  the  bridge  with  a 600 
ohm  line,  the  value  of  these  resistors  should 
be  increased  to  300  ohms  each.  Excessive  heat 
should  not  be  used  in  soldering  either  the  re- 
sistors or  the  diodes  to  ensure  that  their  char- 
acteristics will  not  be  altered  by  application  of 
high  temperatures  over  an  extended  period. 

Testing  the  When  the  instrument  is 

Balanced  Bridge  completed,  a grid-dip  meter 
may  be  coupled  to  the  in- 
put terminals  via  a two  turn  link.  Be  careful 
not  to  pin  the  bridge  meter.  Place  switch  Si  in 
the  "Calibrate”  (open)  position.  Set  the  grid 
dip  meter  in  the  10-Mc.  to  20-Mc.  range  and 
adjust  the  link  and  calibration  control  Ri  for 
full  scale  meter  reading.  A 500  ohm  carbon 
resistor  placed  across  the  output  terminals  of 
the  unit  should  produce  a zero  meter  reading 
when  Si  is  set  to  the  "SWR”  position.  Various 
values  of  resistance  may  now  be  placed  across 
the  meter  terminals  to  obtain  calibration  points 


for  the  meter  scale.  The  ratio  of  the  external 
resistor  to  the  design  value  of  the  bridge  will 
provide  the  SWR  value  for  any  given  meter 
reading.  For  example,  a 1000  ohm  resistor 
has  a ratio  of  1000/500,  and  will  give  an  indi- 
cated SWR  reading  of  2.  A 1500  ohm  resistor 
will  give  an  indicated  reading  of  3,  a 2000 
ohm  resistor  will  provide  a reading  of  4,  and 
so  on.  Before  each  measurement  is  recorded, 
the  calibrate  control  should  be  set  to  a full 
scale  reading  with  Si  open. 

This  simple  system  of  calibration  will  lead 
to  slight  errors  in  calibration  if  the  regulation 
of  the  r-f  source  is  poor.  That  is,  a change  in 
the  external  calibrating  resistance  will  produce 
a varying  load  on  the  r-f  generator  which  could 
easily  cause  a change  in  the  power  applied  to 
the  bridge.  A separate  diode  r-f  voltmeter 
placed  across  the  pickup  loop  will  enable  the 
input  voltage  to  be  held  to  a constant  value 
and  will  provide  somewhat  more  accurate 
bridge  calibration. 

Using  the  The  bridge  is  placed  in  the  trans- 
Bridge  mission  line  and  driven  from  a 
low  powered  source  having  a min- 
imum of  harmonic  content.  Several  measure- 
ments should  be  made  at  various  frequencies 
within  the  range  of  antenna  operation.  The 
seleaor  switch  is  set  to  the  "Cal.”  position  and 
the  "Calibrate”  potentiometer  is  adjusted  for 
full  scale  meter  reading.  The  switch  is  then 
set  to  the  "SWR”  position  and  a reading  is 
taken.  This  reading  and  others  taken  at  various 
frequencies  may  be  plotted  on  a graph  to  pro- 
vide a "SWR  curve”  for  the  particular  antenna 
and  transmission  line.  Antenna  adjustments 
and  line  balancing  operations  may  now  be 
conducted  to  provide  a smooth  SWR  curve. 


737 


738  Test  Equipment 


THE  RADIO 


Figure  22 

THE  ANTENNASCOPE 

The  radiation  resistance  of  r-f  loads  connected 
across  the  output  receptacle  may  be  quickly 
determined  by  a direct  dial  reading.  The  An- 
tennascope  may  be  driven  with  a grid-dip 
oscillator,  covering  r-f  impedance  range  of  5 
to  1000  ohms. 


with  the  point  of  minimum  SWR  occuring  at 
the  chosen  design  frequency  of  the  antenna  in- 
stallation. The  complete  adjustment  and  check- 
out procedure  for  an  antenna  and  transmission 
line  system  is  covered  in  the  Beam  Antenna 
Handbook,  published  by  Radio  Publications, 
Inc.,  Wilton,  Conn. 

34-9  The  Antennascope 

The  Antennascope  is  a modified  SWR 
bridge  in  which  one  leg  of  the  bridge  is  com- 
posed of  a non-inductive  variable  resistor.  This 
resistor  is  calibrated  in  ohms,  and  when  its 
setting  is  equal  to  the  radiation  resistance  of 
the  antenna  under  test  the  bridge  is  in  a bal- 
anced state.  If  a sensitive  voltmeter  is  con- 


r 


SCHEMATIC,  ANTENNASCOPE 

Ri — 7000  ohm  composition  potentiometer  ohm- 
ite  type  AB  or  Allen  Bradley  type  J, 
linear  taper 

Rt — 50  ohm,  1-watt  composition  resistor,  IRC 
type  BT,  or  ohmite  "Little  Devil"  (see 
text) 

M — 0-200  d-c  microammeter 


nected  across  the  bridge,  it  will  indicate  a 
voltage  null  at  bridge  balance.  The  radiation 
resistance  of  the  antenna  may  then  be  read  di- 
rectly from  the  calibrated  resistor  of  the  in- 
strument. 

When  the  antenna  under  test  is  in  a non- 
resonant or  reactive  state,  the  null  indication 
on  the  meter  of  the  Antennascope  will  be  in- 
complete. The  frequency  of  the  exciting  sig- 
nal must  then  be  moved  to  the  resonant  fre- 
quency of  the  antenna  to  obtain  accurate  read- 
ings of  radiation  resistance  from  the  dial  of 
the  instrument. 

A typical  Antennascope  is  shown  in  figures 
22  and  24,  and  the  schematic  is  shown  in  fig- 
ure 23.  A 1000  ohm  non-inductive  carbon  po- 
tentiometer serves  as  the  variable  leg  of  the 
bridge.  The  other  legs  are  composed  of  the  50 
ohm  composition  resistor  and  the  radiation  re- 
sistance of  the  antenna.  If  the  radiation  resis- 
tance of  the  external  load  or  antenna  is  50 
ohms  and  the  potentiometer  is  set  at  mid-scale 
the  bridge  is  in  balance  and  the  diode  volt- 
meter will  read  zero.  If  the  radiation  resistance 
of  the  antenna  is  any  value  other  than  50 
ohms,  the  bridge  may  be  balanced  to  this  new 
value  by  varying  the  position  of  the  potentio- 
meter. Bridge  balance  may  be  obtained  with 
non-reactive  loads  in  the  range  of  5 ohms  to 
1000  ohms  with  this  simple  circuit.  When 
measurements  are  conducted  at  the  resonant 
frequency  of  the  antenna  system  the  radiation 
resistance  of  the  installation  may  be  read  di- 
rectly from  the  calibrated  dial  of  the  Anten- 
nascope. Conversely,  a null  reading  of  the  in- 
strument will  occur  at  the  resonant  frequency, 
which  may  easily  be  found  with  the  aid  of  a 
calibrated  receiver  or  frequency  meter. 


HANDBOOK 


Antennascope  739 


Constructing  the  The  Antennascope  is  buik 
Antennascope  within  a sheet  metal  case 
measuring  3 x 6 x z . 
The  indicating  meter  is  placed  at  the  top  of 
the  case,  and  the  r-f  bridge  occupies  the  lower 
portion  of  the  box.  The  input  and  output  co- 
axial fittings  are  mounted  on  each  side  of  the 
box  and  the  non-inductive  50  ohm  resistor  is 
soldered  between  the  center  terminals  of  the 
receptacles. 

The  calibrating  potentiometer  (Ri)  is 
mounted  upon  a phenolic  plate  placed  over  a 
%-inch  hole  drilled  in  the  front  of  the  box. 
This  reduces  the  capacity  to  ground  of  the  po- 
tentiometer to  a minimum.  Placement  of  the 
small  components  within  the  box  may  be  seen 
in  figure  24.  Care  should  be  taken  to  mount 
the  crystal  diode  at  right  angles  to  the  50  ohm 
resistor  to  reduce  capacity  coupling  between 
the  components. 

The  upper  frequency  limit  of  accuracy  of 
the  Antennascope  is  determined  by  the  as- 
sembly technique.  The  unit  shown  will  work 
with  good  accuracy  to  approximately  100  Me. 
Above  this  frequency,  the  self-inductance  of  the 
leads  prevents  a perfect  null  from  being  ob- 
tained. For  operation  in  the  VHF  region,  it 
would  be  wise  to  rearrange  the  components  to 
reduce  lead  length  to  an  absolute  minimum, 
and  to  use  i/4-inch  copper  strap  for  the  r-f 
leads  instead  of  wire. 

Testing  the  When  the  instrument  is  com- 
Antennascope  pleted,  a grid-dip  meter  may 
be  coupled  to  the  input  recep- 
tacle of  the  Antennascope  by  means  of  a two 
mrn  link.  The  frequency  of  excitation  should 
be  in  the  10  Mc.-20  Me.  region.  Coupling 
should  be  adjusted  to  obtain  a half-scale  read- 
ing of  the  meter.  Various  values  of  1-watt 
composition  resistors  up  to  1000  ohms  ate 
then  plugged  into  the  “output”  coaxial  recep- 
tacle and  the  potentiometer  is  adjusted  for  a 
null  on  the  meter.  The  settings  of  the  poten- 
tiometer may  now  be  calibrated  in  terms  of 
the  load  resistor,  the  null  position  indicating 
the  value  of  the  test  resistor.  A calibrated  scale 
for  the  potentiometer  should  be  made,  as 
shown  in  figure  22. 

Using  the  The  antennascope  may  be 

Antennascope  driven  by  a grid-dip  oscillator 
coupled  to  it  by  a two  turn 
link.  Enough  coupling  should  be  used  to  obtain 
at  least  a % scale  reading  on  the  meter  of 
the  Antennascope  with  no  load  conneaed  to 
the  measuring  terminals.  The  Antennascope 
may  be  considered  to  be  a low  range  r-f  ohm- 


Figure  24 

PLACEMENT  OF  PARTS  WITHIN 
THE  ANTENNASCOPE 
With  the  length  of  leads  shown  this  model 
is  useful  up  to  about  100  Me.  Crystal  diode 
should  be  placed  at  right  angles  to  50  o/im 
composition  resistor. 


meter  and  may  be  employed  to  determine  the 
electrical  length  of  quarter-wave  lines,  surge 
impedance  of  transmission  lines,  and  antenna 
resonance  and  radiation  resistance. 

In  general,  the  measuring  terminals  of  the 
Antennascope  are  connected  in  series  with  the 
load  at  a point  of  maximum  current.  This 
means  the  center  of  a dipole,  or  the  base  of  a 
vertical  14 -wave  ground  plane  antenna.  Exci- 
tation is  supplied  to  the  Antennascope,  and  the 
frequency  of  excitation  and  the  resistance  con- 
trol of  the  Antennascope  are  both  varied  until 
a complete  null  is  obtained  on  the  indicating 
meter  of  the  Antennascope.  The  frequency  of 


740  Test  Equipment 


THE  RADIO 


Figure  25 

SIMPLE  SILICON  CRYSTAL  NOISE 
GENERATOR 


the  source  of  excitation  is  now  the  resonant 
frequency  of  the  load,  and  the  radiation  resist- 
ance of  the  load  may  be  read  upon  the  dial  of 
the  Antennascope. 

On  measurements  on  80  and  40  meters,  it 
might  be  found  that  it  is  impossible  to  obtain 
a complete  null  on  the  Antennascope.  This  is 
usually  caused  by  pickup  of  a nearby  broad- 
cast station,  the  rectified  signal  of  the  broad- 
cast station  obscuring  the  null  indication  on 
the  Antennascope.  This  action  is  only  noticed 
when  antennas  of  large  size  are  being  checked. 

34-10  A Silicon  Crystal 
Noise  Generator 

The  limiting  factor  in  signal  reception  above 
25  Me.  is  usually  the  thermal  noise  generated 
in  the  receiver.  At  any  frequency,  however, 
the  tuned  circuits  of  the  receiver  must  be  ac- 
curately aligned  for  best  signal-to-noise  ratio. 
Circuit  changes  (and  even  alignment  changes) 
in  the  r-f  stages  of  a receiver  may  do  much  to 
either  enhance  or  degrade  the  noise  figure  of 
the  receiver.  It  is  exceedingly  hard  to  deter- 
mine whether  changes  of  either  alignment  or 
circ’itry  are  really  providing  a boost  in  signal- 
to-noise  ratio  of  the  receiver,  or  are  merely  in- 
creasing the  gain  (and  noise)  of  the  unit. 


A SILICON  CRYSTAL  NOISE  GENERATOR 

Figure  26 


A simple  means  of  determining  the  degree 
of  actual  sensitivity  of  a receiver  is  to  inject 
a minute  signal  in  the  input  circuit  and  then 
measure  the  amount  of  this  signal  that  is 
needed  to  overcome  the  inherent  receiver 
noise.  The  less  injected  signal  needed  to  over- 
ride the  receiver  noise  by  a certain,  fixed 
amount,  the  more  sensitive  is  the  receiver. 

A simple  source  of  minute  signal  may  be  ob- 
tained from  a silicon  crystal  diode.  If  a small 
d-c  current  is  passed  through  a silicon  crystal 
in  the  direction  of  highest  resistance,  a small 
but  constant  r-f  noise  (or  hiss)  is  generated. 
The  voltage  necessary  to  generate  this  noise 
may  be  obtained  from  a few  flashlight  cells. 
The  generator  is  a broad  band  device  and  re- 
quires no  mning.  If  built  with  short  leads,  it 
may  be  employed  for  receiver  measurements 
well  above  150  Me.  The  noise  generator  should 
be  used  for  comparative  measurements  only, 
since  calibration  against  a high  quality  com- 
mercial noise  generator  is  necessary  for  ab- 
solute measurements. 

A PracHcol  Shown  in  figure  25  is  a sim- 

Noise  Generot-or  pie  silicon  crystal  noise 
generator.  The  schematic  of 
this  unit  is  illustrated  in  figure  26.  The  1N21 
crystal  and  .001  gfd.  ceramic  capacitor  are 
connected  in  series  directly  across  the  output 
terminals  of  the  instrument.  Three  small  flash- 
light batteries  are  wired  in  series  and  mounted 
inside  the  case,  along  with  the  0-2  d-c  milliam- 
meter  and  the  noise  level  potentiometer. 

To  prevent  heat  damage  to  the  1N21  crystal 
during  the  soldering  process,  the  crystal  should 
be  held  with  a damp  rag,  and  the  connections 
soldered  to  it  quickly  with  a very  hot  iron. 
Across  the  terminals  (and  in  parallel  with  the 
equipment  to  be  attached  to  the  generator)  is 
a 1-watt  carbon  resistor  whose  resistance  is 
equal  to  the  impedance  level  at  which  meas- 
urements are  to  be  made.  This  will  usually  be 
either  50  or  300  ohms.  If  the  noise  generator 
is  to  be  used  at  one  impedance  level  only,  this 


HANDBOOK 


Noise  Generator  741 


NOISE 

GENERATOR 

m @ 

1 — ► 

RECEIVER 

-o 

o- 

ANT. 

SPKR 

-0 

O 

RESISTOR 

TEST  SET-UP  FOR  NOISE  GENERATOR 


Figure  27 


resistor  may  be  mounted  permanently  inside 
of  the  case. 

Using  the  The  test  setup  for  use  of 

Noise  Generator  the  noise  generator  is  shown 
in  figure  27.  The  noise  gen- 


erator is  connected  to  the  antenna  terminals 
of  the  receiver  under  test.  The  receiver  is 
turned  on,  the  a.v.c.  turned  off,  and  the  r-f 
gain  control  placed  full  on.  The  audio  volume 
control  is  adjusted  until  the  output  meter  ad- 
vances to  one-quarter  scale.  This  reading  is 
the  basic  receiver  noise.  The  noise  generator 
is  turned  on,  and  the  noise  level  potentiometer 
adjusted  until  the  noise  output  voltage  of  the 
receiver  is  doubled.  The  more  resistance  in 
the  diode  circuit,  the  better  is  the  signal-to- 
noise  ratio  of  the  receiver  under  test.  The  r-f 
circuits  of  the  receiver  may  be  aligned  for 
maximum  signal-to-noise  ratio  with  the  noise 
generator  by  aligning  for  a 2/1  noise  ratio  at 
minimum  diode  current. 


CHAPTER  THIRTY-FIVE 


Radio  Mathematics 
and  Calculations 


Radiomen  often  have  occasion  to  cal- 
culate sizes  and  values  of  required  parts.  This 
requires  some  knowledge  of  mathematics.  The 
following  pages  contain  a review  of  those  parts 
of  mathematics  necessary  to  understand  and 
apply  the  information  contained  in  this  book. 
It  is  assumed  that  the  reader  has  had  some 
mathematical  training;  this  chapter  is  not  in- 
tended to  teach  those  who  have  never  learned 
anything  of  the  subject. 

Fortunately  only  a knowledge  of  fundamen- 
tals is  necessary,  although  this  knowledge  must 
include  several  branches  of  the  subject.  Fortu- 
nately, too,  the  majority  of  practical  applica- 
tions in  radio  work  reduce  to  the  solution  of 
equations  or  formulas  or  the  interpretation  of 
graphs. 

Arithmetic 

Notation  of  In  writing  numbers  in  the  Ara- 
Numbers  bic  system  we  employ  ten  dif- 

ferent symbols,  digits,  or  fig- 
ures: 1,  2,  3,  4,  3,  6,  7,  8,  9,  and  0,  and  place 
them  in  a definite  sequence.  If  there  is  more 
than  one  figure  in  the  number  the  position  of 
each  figure  or  digit  is  as  important  in  deter- 
mining its  value  as  is  the  digit  itself.  When  we 
deal  with  whole  numbers  the  righthandmost 
digit  represents  units,  the  next  to  the  left  rep- 
resents tens,  the  next  hundreds,  the  next  thou- 
sands, from  which  we  derive  the  rule  that  ev- 
ery time  a digit  is  placed  one  space  further  to 
the  left  its  value  is  multiplied  by  ten. 

8 14  3 

thousands  hundreds  tens  units 

It  will  be  seen  that  any  number  is  actually  a 
sum.  In  the  example  given  above  it  is  the  sum 
of  eight  thousands,  plus  one  hundred,  plus 


four  tens,  plus  three  units,  which  could  be 
written  as  follows: 

8 thousands  ( 1 0 x 1 0 x 10) 

1 hundreds  ( 1 0 x 10) 

4 tens 
3 units 


8143 

The  number  in  the  units  position  is  some- 
times referred  to  as  a first  order  number,  that 
in  the  tens  position  is  of  the  second  order,  that 
in  the  hundreds  position  the  third  order,  etc. 

The  idea  of  letting  the  position  of  the  sym- 
bol denote  its  value  is  an  outcome  of  the  aba- 
cus. The  abacus  had  only  a limited  num- 
ber of  wires  with  beads,  but  it  soon  became 
apparent  that  the  quantity  of  symbols  might 
be  continued  indefinitely  towards  the  left, 
each  further  space  multiplying  the  digit’s 
value  by  ten.  Thus  any  quantity,  however 
large,  may  readily  be  indicated. 

It  has  become  customary  for  ease  of  reading 
to  divide  large  numbers  into  groups  of  three 
digits,  separating  them  by  commas. 

6,000,000  rather  than  6000000 

Our  system  of  notation  then  is  characterized 
by  two  things:  the  use  of  positions  to  indicate 
the  value  of  each  symbol,  and  the  use  of  ten 
symbols,  from  which  we  derive  the  name  dec- 
imal system. 

Retaining  the  same  use  of  positions,  we 
might  have  used  a different  number  of  sym- 
bols, and  displacing  a symbol  one  place  to  the 
left  might  multiply  its  value  by  any  other  fac- 
tor such  as  2,  6 or  12.  Such  other  systems  have 
been  in  use  in  history,  but  will  not  be  discussed 
here.  There  are  also  systems  in  which  displac- 
ing a symbol  to  the  left  multiplies  its  value  by 


742 


Decimal  Fractions  743 


0 0.5  I 2 3 4 


0.7 


varying  factors  in  accordance  with  complicat- 
ed rules.  The  English  system  of  measurements 
is  such  an  inconsistent  and  inferior  system. 

Decimal  Fractions  Since  we  can  extend  a 
number  indefinitely  to 
the  left  to  make  it  bigger,  it  is  a logical  step  to 
extend  it  towards  the  right  to  make  it  smaller. 
Numbers  smaller  than  unity  ate  jractions  and 
if  a displacement  one  position  to  the  right  di- 
vides its  value  by  ten,  then  the  number  is  re- 
ferred to  as  a decimal  jraction.  Thus  a digit 
to  the  tight  of  the  units  column  indicates  the 
number  of  tenths,  the  second  digit  to  the  right 
represents  the  number  of  hundredths,  the  third, 
the  number  of  thousandths,  etc.  Some  distin- 
guishing mark  must  be  used  to  divide  unit  from 
tenths  so  that  one  may  properly  evaluate  each 
symbol.  This  mark  is  the  decimal  point. 

A decimal  fraction  like  jour-tenths  may  be 
written  .4  or  0.4  as  desired,  the  latter  probably 


being  the  clearer.  Every  time  a digit  is  placed 
one  space  further  to  the  right  it  represents  a 
ten  times  smaller  part.  This  is  illustrated  in 
Figure  1,  where  each  large  division  represents 
a unit;  each  unit  may  be  divided  into  ten  parts 
although  in  the  drawing  we  have  only  so  di- 
vided the  first  part.  The  length  ah  is  equal  to 
seven  of  these  tenth  parts  and  is  written  as  0.7. 

The  next  smaller  divisions,  which  should  be 
written  in  the  second  column  to  the  right  of 
the  decimal  point,  are  each  one-tenth  of  the 
small  division,  or  one  one-hundredth  each. 
They  are  so  small  that  we  can  only  show  them 
by  imagining  a magnifying  glass  to  look  at 
them,  as  in  Figure  1.  Six  of  these  divisions  is 
to  be  written  as  0.06  (six  hundredths).  We 
need  a microscope  to  see  the  next  smaller  divi- 
sion, that  is  those  in  the  third  place,  which  will 
be  a tenth  of  one  one-hundredth,  or  a thou- 
sandth; four  such  divisions  would  be  written 
as  0.004  (four  thousandths). 


Figure  2. 

IN  THIS  ILLUSTRATION  FRACTIONAL  PORTIONS  ARE  REPRESENTED  IN  THE 
FORM  OF  RECTANGLES  RATHER  THAN  LINEARLY. 

ABCS  = I.OrGffD  = 0.1;  KJEH  = 0.01;  eoeh  small  section  within  KJIH  equals  0.001 


744  Radio  Mathematics  and  Calculations  THE  RADIO 


It  should  not  be  thought  that  such  numbers 

are  merely  of  academic  interest  for  very  small 

quantities  are  common  in  radio  work. 

Possibly  the  conception  of  fractions  may  be 
clearer  to  some  students  by  representing  it  in 
the  form  of  rectangles  rather  than  linearly 
(see  Figure  2). 

Addition  When  two  or  more  numbers  are 
to  be  added  we  sometimes  write 
them  horizontally  with  the  plus  sign  between 
them.  + is  the  sign  or  operator  indicating  ad- 
dition. Thus  if  7 and  12  are  to  be  added  to- 
gether we  may  write  7-1-12  = 19. 

But  if  larger  or  more  numbers  are  to  be  added 
together  they  are  almost  invariably  written  one 
under  another  in  such  a position  that  the  deci- 
mal points  fall  in  a vertical  line.  If  a number 
has  no  decimal  point,  it  is  still  considered  as 
being  just  to  the  right  of  the  units  figure;  such 
a number  is  a whole  number  or  integer.  Ex- 
amples: 


654  0.654  654 

32  3.2  32 

53041  53.041  5304.1 


53727  56.895  5990.1 


The  result  obtained  by  adding  numbers  is 
called  the  .turn. 

Subtraction  Subtraction  is  the  reverse  of 
addition.  Its  operator  is  — (the 
minii.<  sign).  The  number  to  be  subtracted  is 
called  the  minuend,  the  number  from  which  it 
is  subtracted  is  the  subtrahend,  and  the  result 
is  called  the  remainder. 

subtrahend 
— minuend 


remainder 

Examples: 

65.4  65.4 

-32  -32.21 


33.4  33.19 

Multiplication  When  numbers  are  to  be  mul- 
tiplied together  we  use  the  X , 
which  is  known  as  the  multiplication  or  the 
times  sign.  The  number  to  be  multiplied  is 
known  as  the  multiplicand  and  that  by  which 
it  is  to  be  multiplied  is  the  multiplier,  which 
may  be  written  in  words  as  follows: 

multiplicand 
X multiplier 


partial  product 
partial  product 


product 


The  result  of  the  operation  is  called  the 

product. 

From  the  examples  to  follow  it  will  be  obvi- 
ous that  there  are  as  many  partial  products 
as  there  are  digits  in  the  multiplier.  In  the  fol- 
lowing examples  note  that  the  righthandmost 
digit  of  each  partial  product  is  placed  one 
space  farther  to  the  left  than  the  previous  one. 


834 
X 26 


5004 

1668 


21684 


834 
X 206 


5004 

000 

1668 


171804 


In  the  second  example  above  it  will  be  seen 
that  the  inclusion  of  the  second  partial  prod- 
uct was  unnecessary;  whenever  the  multiplier 
contains  a cipher  (zero)  the  next  partial  prod- 
uct should  be  moved  an  additional  space  to 
the  left. 

Numbers  containing  decimal  fractions  may 
first  be  multiplied  exactly  as  if  the  decimal 
point  did  not  occur  in  the  numbers  at  all;  the 
position  of  the  decimal  point  in  the  product  is 
determined  after  all  operations  have  been  com- 
pleted. It  must  be  so  positioned  in  the  product 
that  the  number  of  digits  to  its  right  is  equal 
to  the  number  of  decimal  places  in  the  multi- 
plicand plus  the  number  of  decimal  places  in 
the  multiplier. 

This  rule  should  be  well  understood  since 
many  radio  calculations  contain  quantities 
which  involve  very  small  decimal  fractions.  In 
the  examples  which  follow  the  explanatory 
notations  "2  places,”  etc.,  are  not  actually 
written  down  since  it  is  comparatively  easy  to 
determine  the  decimal  point’s  proper  location 
mentally. 

5.43  2 places 

X0.72  2 places 


1086 
3 801 


3.9096  2 -f- 2 = 4 places 

0.04  2 places 

X 0.003  3 pi  aces 

0.00012  2 + 3-5  places 

Division  Division  is  the  reverse  of  multi- 
plication. Its  operator  is  the  =, 
which  is  called  the  division  sign.  It  is  also  com- 
mon to  indicate  division  by  the  use  of  the  frac- 
tion bar  (/)  or  by  writing  one  number  over 
the  other.  The  number  which  is  to  be  divided 
is  called  the  dividend  and  is  written  before 
the  division  sign  or  fraction  bar  or  over  the 
horizontal  line  indicating  a fraction.  The  num- 


HANDBOOK 


Division  745 


ber  by  which  the  dividend  is  to  be  divided  is 
called  the  divisor  and  follows  the  division 
sign  or  fraction  bar  or  comes  under  the  hori- 
zontal line  of  the  fraction.  The  answer  or 
result  is  called  the  quotient. 

quotient 
divisor  ) dividend 

or 

dividend  ^divisor = quotient 

or 

dividend 


divisor 

Examples: 

126 

834  ) 105084 
834 


2168 

1668 


49 

49 ) 2436 
196 


476 

441 


Another  example:  Divide  0.000325  by  0.017. 
Here  we  must  move  the  decimal  point  three 
places  to  the  right  in  both  dividend  and  di- 
visor. 

0.019 
17)  0.325 
17 

155 

153 

2 


In  a case  where  the  dividend  has  fewer  deci- 
mals than  the  divisor  the  same  rules  still  may 
be  applied  by  adding  ciphers.  For  example  to 
divide  0.49  by  0.006  we  must  move  the 
decimal  point  three  places  to  the  right.  The 
0.49  now  becomes  490  and  we  write: 


§2^ 

6 7190 

48 


5004  35  remainder 

5004 


10 

6 


Note  that  one  number  often  fails  to  divide 
into  another  evenly.  Hence  there  is  often  a 
quantity  left  over  called  the  remainder. 

The  rules  for  placing  the  decimal  point 
are  the  reverse  of  those  for  multiplication. 
The  number  of  decimal  places  in  the  quotient 
is  equal  to  the  difference  between  the  number 
of  decimal  places  in  the  dividend  and  that  in 
the  divisor.  It  is  often  simpler  and  clearer 
to  remove  the  decimal  point  entirely  from  the 
divisor  by  multiplying  both  dividend  and  di- 
visor by  the  necessary  factor;  that  is  we  move 
the  decimal  point  in  the  divisor  as  many  places 
to  the  right  as  is  necessary  to  make  it  a whole 
number  and  then  we  move  the  decimal  point 
in  the  dividend  exactly  the  same  number  of 
places  to  the  right  regardless  of  whether  this 
makes  the  dividend  a whole  number  or  not. 
When  this  has  been  done  the  decimal  point 
in  the  quotient  will  automatically  come  di- 
rectly above  that  in  the  dividend  as  shown  in 
the  following  example. 

Example:  Divide  10.5084  by  8.34.  Move  the 
decimal  point  of  both  dividend  and  divisor 
two  places  to  the  right. 

1.26 

834')  1050.84 
834 

2168 

1668 

5004 

5004 


4 

When  the  division  shows  a remainder  it  is 
sometimes  necessary  to  continue  the  work  so 
as  to  obtain  more  figures.  In  that  case  ciphers 
may  be  annexed  to  the  dividend,  brought  down 
to  the  remainder,  and  the  division  continued 
as  long  as  may  be  necessary;  be  sure  to  place 
a decimal  point  in  the  dividend  before  the 
ciphers  are  annexed  if  the  dividend  does  not 
already  contain  a decimal  point.  For  ex- 
ample: 

80.33 
6 ) 482.00 
48 

20 

18 

20 

18 

~Z 

This  operation  is  not  very  often  required 
in  radio  work  since  the  accuracy  of  the  mea- 
surements from  which  our  problems  start 
seldom  justifies  the  use  of  more  than  three 
significant  figures.  This  point  will  be  cov- 
ered further  later  in  this  chapter. 

Fractions  Quantities  of  less  than  one 
(unity)  are  called  fractions.  They 
may  be  expressed  by  decimal  notation  as  we 
have  seen,  or  they  may  be  expressed  as  vulgar 
fractions.  Examples  of  vulgar  fractions: 


746  Radio  Mathematics  and  Calculations  THE  RADIO 


numerator  3 6 1 

denominator  4 7 5 

The  upper  position  of  a vulgar  fraction  is 
called  the  numerator  and  the  lower  position 
the  denominator.  When  the  numerator  is  the 
smaller  of  the  two,  the  fraction  is  called  a 
proper  fraction;  the  examples  of  vulgar  frac- 
tions given  above  are  proper  vulgar  fractions. 
When  the  numerator  is  the  larger,  the  ex- 
pression is  an  improper  fraction,  which  can 
be  reduced  to  an  integer  or  whole  number 
with  a proper  fraction,  the  whole  being  called 
a mixed  number.  In  the  following  examples 
improper  fractions  have  been  reduced  to 
their  corresponding  mixed  numbers. 


Adding  or  Subtracting  Except  when  the 

Fractions  fractions  are  very 

simple  it  will  usual- 
ly be  found  much  easier  to  add  and  subtract 
fractions  in  the  form  of  decimals.  This  rule 
likewise  applies  for  practically  all  other  oper- 
ations with  fractions.  However,  it  is  occa- 
sionally necessary  to  perform  various  opera- 
tions with  vulgar  fractions  and  the  rules 
should  be  understood. 

When  adding  or  subtracting  such  fractions 
the  denominators  must  be  made  equal.  This 
may  be  done  by  multiplying  both  numerator 
and  denominator  of  the  first  fraction  by  the 
denominator  of  the  other  fraction,  after 
which  we  multiply  the  numerator  and  de- 
nominator of  the  second  fraction  by  the  de- 
nominator of  the  first  fraction.  This  sounds 
more  complicated  than  it  usually  proves  in 
practice,  as  the  following  examples  will  show. 

J_4.J 5- 

2^3  — L2x3T^3x2j—  6 + 6 — 6 

3 J r 3x5  2x4~|  _ 15  _i 7_ 

4 5 — |_4x5  5x4j  — 20~20  — 20 

Except  in  problems  involving  large  numbers 
the  step  shown  in  brackets  above  is  usually 
done  in  the  head  and  is  not  written  down. 

Although  in  the  examples  shown  above  we 
have  used  proper  fractions,  it  is  obvious  that 
the  same  procedure  applies  with  improper 
fractions.  In  the  case  of  problems  involving 
mixed  numbers  it  is  necessary  first  to  convert 
them  into  improper  fractions.  Example: 

,3 2x7  -P  3 _ 17 

^ 7 — 7 — 7 

The  numerator  of  the  improper  fraction  is 
equal  to  the  whole  number  multiplied  by  the 
denominator  of  the  original  fraction,  to  which 


the  numerator  is  added.  That  is  in  the  above 
example  we  multiply  2 by  7 and  then  add  3 
to  obtain  17  for  the  numerator.  The  denomi- 
nator is  the  same  as  is  the  denominator  of 
the  original  fraction.  In  the  following  ex- 
ample we  have  added  two  mixed  numbers. 


2-I-  + 


3 _ 17  , 15  _ ri7x4 

4 — 7 + 4 — |_  7x4 


+ 


15x7 

4x7 


] 


— 28  + 28  — 28  — ° 28 


Multiplying  All  vulgar  fractions  are  multi- 
Fractions  plied  by  multiplying  the  nu- 
merators together  and  the  de- 
nominators together,  as  shown  in  the  follow- 
ing example: 

-3  V 2 -f  1 i 

4 -^5  — L4x5j  — 20  — 10 

As  above,  the  step  indicated  in  brackets  is 
usually  not  written  down  since  it  may  easily 
be  performed  mentally.  As  with  addition  and 
subtraction  any  mixed  numbers  should  be 
first  reduced  to  improper  fractions  as  shown 
in  the  following  example: 

J_v  dJ L- --11^ 

2j  A.  ■»  j — 23  A 3 — 69  — 23 

Division  of  Fractions  may  be  most  easily 
Fractions  divided  by  inverting  the  di- 
visor and  then  multiplying. 

Example: 

^ . _? V-? ®- 

5 • 4 — 5-^3  - 15 


In  the  above  example  it  will  be  seen  that  to 
divide  by  % is  exactly  the  same  thing  as  to 
multiply  by  4/3.  Actual  division  of  fractions 
is  a rather  rare  operation  and  if  necessary  is 
usually  postponed  until  the  final  answer  is  se- 
cured when  it  is  often  desired  to  reduce  the 
resulting  vulgar  fraction  to  a decimal  frac- 
tion by  division.  It  is  more  common  and 
usually  results  in  least  overall  work  to  re- 
duce vulgar  fractions  to  decimals  at  the  be- 
ginning of  a problem.  Examples: 

-|-=  0.375  -^=  0.15625 

0.15625 
32  ) 5.00000 

H 
1 80 
1 60 
200 
192 
80 
64 
160 
160 


HANDBOOK 


Division  of  Fractions  747 


It  will  be  obvious  that  many  vulgar  fractions 
cannot  be  reduced  to  exact  decimal  equiva- 
lents. This  fact  need  not  worry  us,  however, 
since  the  degree  of  equivalence  can  always  be 
as  much  as  the  data  warrants.  For  instance, 
if  we  know  that  one-third  of  an  ampere  is 
flowing  in  a given  circuit,  this  can  be  written 
as  0.33.^  amperes.  This  is  not  the  exact 
equivalent  of  1/3  but  is  close  enough  since  it 
shows  the  value  to  the  nearest  thousandth  of 
an  ampere  and  it  is  probable  that  the  meter 
from  which  we  secured  our  original  data  was 
not  accurate  to  the  nearest  thousandth  of  an 
ampere. 

Thus  in  converting  vulgar  fractions  to  a 
decimal  we  unhesitatingly  stop  when  we  have 
reached  the  number  of  significant  figures  war- 
ranted by  our  original  data,  which  is  very 
seldom  more  than  three  places  (see  section 
Significant  Figures  later  in  this  chapter). 

When  the  denominator  of  a vulgar  fraction 
contains  only  the  factors  2 or  5,  division  can 
be  brought  to  a finish  and  there  will  be  no 
remainder,  as  shown  in  the  examples  above. 

When  the  denominator  has  other  factors 
such  as  3,  7,  11,  etc.,  the  division  will  seldom 
come  out  even  no  matter  how  long  it  is  con- 
tinued but,  as  previously  stated,  this  is  of 
no  consequence  in  practical  work  since  it  may 
be  carried  to  whatever  degree  of  accuracy  is 
necessary.  The  digits  in  the  quotient  will 
usually  repeat  either  singly  or  in  groups,  al- 
though there  may  first  occur  one  or  more 
digits  which  do  not  repeat.  Such  fractions 
are  known  as  repeating  fractions.  They  are 
sometimes  indicated  by  an  oblique  line  (frac- 
tion bar)  through  the  digit  which  repeats,  or 
through  the  first  and  last  digits  of  a repeating 
group.  Example; 

^ = 0.3333  =0.^ 

~ = 0.142857142857  = 0.^4285^ 

The  foregoing  examples  contained  only  re- 
peating digits.  In  the  following  example  a 
non-repeating  digit  precedes  the  repeating 
digit: 

^ = 0.2333  = 0.2^ 

While  repeating  decimal  fractions  can  be 
converted  into  their  vulgar  fraction  equiva- 
lents, this  is  seldom  necessary  in  practical 
work  and  the  rules  will  be  omitted  here. 

Powers  and  When  a number  is  to  be  mul- 

Roots  tiplied  by  itself  we  say  that 

it  is  to  be  squared  or  to  be 
raised  to  the  second  power.  When  it  is  to  be 
multipled  by  itself  once  again,  we  say  that 
it  is  cubed  or  raised  to  the  third  power. 


In  general  terms,  when  a number  is  to  be 
multipled  by  itself  we  speak  of  raising  to  a 
power  or  involution;  the  number  of  times 
which  the  number  is  to  be  multiplied  by  it- 
self is  called  the  order  of  the  power.  The 
standard  notation  requires  that  the  order  of 
the  power  be  indicated  by  a small  number 
written  after  the  number  and  above  the  line, 
called  the  exponent.  Examples: 

2‘  z=  2 X 2,  or  2 squared,  or  the  second 
power  of  2 

2^  = 2 X 2 X 2,  or  2 cubed,  or  the  third 
power  of  2 

2'  = 2 X 2 X 2 X 2,  or  the  fourth  pow- 
er of  2 

Sometimes  it  is  necessary  to  perform  the 
reverse  of  this  operation,  that  is,  it  may  be 
necessary,  for  instance,  to  find  that  number 
which  multiplied  by  itself  will  give  a product 
of  nine.  The  answer  is  of  course  3.  This 
process  is  known  as  extracting  the  root  or 
evolution.  The  particular  example  which  is 
cited  would  be  written: 

\/9  = 3 

The  sign  for  extracting  the  root  is 
which  is  known  as  the  radical  sign;  the  order 
of  the  root  is  indicated  by  a small  number 
above  the  radical  as  in  which  would  mean 
the  fourth  root;  this  number  is  called  the 
index.  When  the  radical  bears  no  index,  the 
square  or  second  root  is  intended. 

Restricting  our  attention  for  the  moment 
to  square  root,  we  know  that  2 is  the  square 
root  of  4,  and  3 is  the  square  root  of  9.  If 
we  want  the  square  root  of  a number  between 
3 and  9,  such  as  the  square  root  of  5,  it  is 
obvious  that  it  must  lie  between  2 and  3.  In 
general  the  square  root  of  such  a number  can- 
not be  exactly  expressed  either  by  a vulgar 
fraction  or  a decimal  fraction.  However,  the 
square  root  can  be  carried  out  decimally  as 
far  as  may  be  necessary  for  sufficient  accur- 
acy. In  general  such  a decimal  fraction  will 
contain  a never-ending  series  of  digits  with- 
out repeating  groups.  Such  a number  is  an 
irrational  number,  such  as 

t/5  = 2.2361  

The  extraction  of  roots  is  usually  done  by 
tables  or  logarithms  the  use  of  which  will 
be  described  later.  There  are  longhand  meth- 
ods of  extracting  various  roots,  but  we  shall 
give  only  that  for  extracting  the  square  root 
since  the  others  become  so  tedious  as  to  make 
other  methods  almost  invariably  preferable. 
Even  the  longhand  method  for  extracting  the 
square  root  will  usually  be  used  only  if  loga- 


748  Radio  Mathematics  and  Calculations  THE  RADIO 


rithm  tables,  slide  rule,  or  table  of  roots  are 

not  handy. 

Extracting  the  First  divide  the  number  the 
Square  Root  root  of  which  is  to  be  ex- 
tracted into  groups  of  two 
digits  starting  at  the  decimal  point  and  going 
in  both  directions.  If  the  lefthandmost  group 
proves  to  have  only  one  digit  instead  of  two, 
no  harm  will  be  done.  The  righthandmost 
group  may  be  made  to  have  two  digits  by 
annexing  a zero  if  necessary.  For  example, 
let  it  be  required  to  find  the  square  root  of 
5678.91.  This  is  to  be  divided  off  as  follows: 

Vse'  78.91 

The  mark  used  to  divide  the  groups  may 
be  anything  convenient,  although  the  prime- 
sign  (')  is  most  commonly  used  for  the 
purpose. 

Next  find  the  largest  square  which  is  con- 
tained in  the  first  group,  in  this  case  56.  The 
largest  square  is  obviously  49,  the  square  of  7. 
Place  the  7 above  the  first  group  of  the  num- 
ber whose  root  is  to  be  extracted,  which  is 
sometimes  called  the  dividend  from  analogy 
to  ordinary  division.  Place  the  square  of  this 
figure,  that  is  49,  under  the  first  group,  56, 
and  subtract  leaving  a remainder  of  7. 

7 

/56'  78.91 
49 

7 

Bring  down  the  next  group  and  annex  it  to 
the  remainder  so  that  we  have  778.  Now  to 
the  left  of  this  quantity  write  down  twice  the 
root  so  far  found  (2  X 7 or  14  in  this  ex- 
ample), annex  a cipher  as  a trial  divisor,  and 
see  how  many  times  the  result  is  contained 
in  778.  In  our  example  140  will  go  into  778 
5 times.  Replace  the  cipher  with  a 5,  and 
multiply  the  resulting  145  by  5 to  give  725. 
Place  the  5 directly  above  the  second  group 
in  the  dividend  and  then  subtract  the  725 
from  778. 

7 5 

\f56'  78.91 
49 

140  7 78 

145  X 5 = 7 25 

~53 

The  next  step  is  an  exact  repetition  of  the 
previous  step.  Bring  down  the  third  group 
and  annex  it  to  the  remainder  of  53,  giving 
5391.  Write  down  twice  the  root  already 


found  and  annex  the  cipher  (2  x 75  or  150 
plus  the  cipher,  which  will  give  1500).  1500 
will  go  into  5391  3 times.  Replace  the  last 
cipher  with  a three  and  multiply  1503  by  3 to 
give  4509.  Place  3 above  the  third  group. 
Subtract  to  find  the  remainder  of  882.  The 
quotient  75.3  which  has  been  found  so  far  is 
not  the  exact  square  root  which  was  desired; 
in  most  cases  it  will  be  sufficiently  accurate. 
However,  if  greater  accuracy  is  desired  groups 
of  two  ciphers  can  be  brought  down  and  the 
process  carried  on  as  long  as  necessary. 

7 5.  3 
756'  78.91 
49 

140  7 78 

145  X 5 = 7 25 

1500  53  91 

1 503  X 3 = 45  09 


8 82 

Each  digit  of  the  root  should  be  placed  di- 
rectly above  the  group  of  the  dividend  from 
which  it  was  derived;  if  this  is  done  the 
decimal  point  of  the  root  will  come  directly 
above  the  decimal  point  of  the  dividend. 

Sometimes  the  remainder  after  a square 
has  been  subtracted  (such  as  the  1 in  the  fol- 
lowing example)  will  not  be  sufficiently  large 
to  contain  twice  the  root  already  found  even 
after  the  next  group  of  figures  has  been 
brought  down.  In  this  case  we  write  a cipher 
above  the  group  just  brought  down  and  bring 
down  another  group. 

7.  0 8 2 

750.16'  00'  00 

49 


1400 

1 16 

00 

1408 

X 

8 = 

1 12 

64 

14160 

3 

36 

00 

14162 

X 

2 = 

2 

83 

24 

52 

76 

In  the  above  example  the  amount  116  was  not 
sufficient  to  contain  twice  the  root  already 
found  with  a cipher  annexed  to  it;  that  is, 
it  was  not  sufficient  to  contain  140.  There- 
fore we  write  a zero  above  16  and  bring  down 
the  next  group,  which  in  this  example  is  a 
pair  of  ciphers. 

Order  of  One  frequently  encounters  prob- 
Operations  lems  in  which  several  of  the  fun- 
damental operations  of  arithme- 
tic which  have  been  described  are  to  be  per- 
formed. The  order  in  which  these  operations 


HANDBOOK 


Order  of  Operations  749 


must  be  performed  is  important.  First  al!  pow- 
ers and  roots  should  be  calculated;  multipli- 
cation and  division  come  next;  adding  and 
subtraction  come  last.  In  the  example 

2 -f  3 X 

we  must  first  square  the  4 to  get  16;  then  we 
multiply  16  by  3,  making  48,  and  to  the 
product  we  add  2,  giving  a result  of  50. 

If  a different  order  of  operations  were  fol- 
lowed, a different  result  would  be  obtained. 
For  instance,  if  we  add  2 to  3 we  would  ob- 
tain 5,  and  then  multiplying  this  by  the  square 
of  4 or  16,  we  would  obtain  a result  of  80, 
which  is  incorrect. 

In  more  complicated  forms  such  as  frac- 
tions whose  numerators  and  denominators  may 
both  be  in  complicated  forms,  the  numerator 
and  denominator  are  first  found  separately 
before  the  division  is  made,  such  as  in  the 
following  example: 

3X4  + 5xa_  12-plO  __  22  _ , 
2X3-(-2-t-3— 6-|-2-f3~  11  — ^ 

Problems  of  this  type  are  very  common  in 
dealing  with  circuits  containing  several  in- 
ductances, capacities,  or  resistances. 

The  order  of  operations  specified  above  does 
not  always  meet  all  possible  conditions;  if  a 
series  of  operations  should  be  performed  in  a 
different  order,  this  is  always  indicated  by 
parentheses  or  brackets,  for  example: 

2-1-3X4^  = 2 + 3X16  = 2 + 48  = 50 
(2  + 3)  X 4*  = 5 X 4’  = 5 X 16  = 80 
2 + (3  X 4)"=  2 + 12"  = 2 + 144  = 146 

In  connection  with  the  radical  sign,  brackets 
may  be  used  or  the  "hat”  of  the  radical  may 
be  extended  over  the  entire  quantity  whose 
root  is  to  be  extracted.  Example: 

V4  + 5=  /4  + 5 = 2+  5 = 7 

V (4  + 5)  = V4  + 5 = /9  = 3 

It  is  recommended  that  the  radical  always  be 
extended  over  the  quantity  whose  root  is  to  be 
extracted  to  avoid  any  ambiguity. 

Cancellation  In  a fraction  in  which  the 
numerator  and  denominator 
consist  of  several  factors  to  be  multiplied,  con- 
siderable labor  can  often  be  saved  if  it  is 
found  that  the  same  factor  occurs  in  both 
numerator  and  denominator.  These  factors 
cancel  each  other  and  can  be  removed.  Ex- 
ample; 


In  the  foregoing  example  it  is  obvious  that  the 
3 in  the  numerator  goes  into  the  6 in  the  de- 
nominator twice.  We  may  thus  cross  out 
the  three  and  replace  the  6 by  a 2.  The  2 
which  we  have  just  placed  in  the  denominator 
cancels  the  2 in  the  numerator.  Next  the  5 
in  the  denominator  will  go  into  the  25  in  the 
numerator  leaving  a result  of  5.  Now  we 
have  left  only  a 5 in  the  numerator  and  a 7 
in  the  denominator,  so  our  final  result  is  5/7. 
If  we  had  multiplied  2 x 3 X 25  to  obtain 
150  and  then  had  divided  this  by  6 X 5 X 7 
or  210,  we  would  have  obtained  the  same  re- 
sult but,  with  considerably  more  work. 

Algebra 

Algebra  is  not  a separate  branch  of  mathe- 
matics but  is  merely  a form  of  generalized 
arithmetic  in  which  letters  of  the  alphabet  and 
occasional  other  symbols  are  substituted  for 
numbers,  from  which  it  is  often  referred  to  as 
literal  notation.  It  is  simply  a shorthand  meth- 
od of  writing  operations  which  could  be  spelled 
out. 

The  laws  of  most  common  electrical  phe- 
nomena and  circuits  (including  of  course  ra- 
dio phenomena  and  circuits)  lend  themselves 
particularly  well  to  representation  by  literal  no- 
tation and  solution  by  algebraic  equations  or 
formulas. 

While  we  may  write  a particular  problem  in 
Ohm's  Law  as  an  ordinary  division  or  multi- 
plication, the  general  statement  of  all  such 
problems  calls  for  the  replacement  of  the  num- 
bers by  symbols.  We  might  be  explicit  and 
write  out  the  names  of  the  units  and  use  these 
names  as  symbols: 

volts  = amperes  X ohms 

Such  a procedure  becomes  too  clumsy  when 
the  expression  is  more  involved  and  would  be 
unusually  cumbersome  if  any  operations  like 
multiplication  were  required.  Therefore  as  a 
short  way  of  writing  these  generalized  rela- 
tions the  numbers  are  represented  by  letters. 
Ohm’s  Law  then  becomes 

E = I X R 

In  the  statement  of  any  particular  problem 
the  significance  of  the  letters  is  usually  indi- 
cated directly  below  the  equation  or  formula 
using  them  unless  there  can  be  no  ambiguity. 
Thus  the  above  form  of  Ohm’s  Law  would  be 
more  completely  written  as: 


750  Radio  Mathematics  and  Calculations  THE  RADIO 


E = I X R 

where  E = e.m.f.  in  volts 

I = current  in  amperes 
R = resistance  in  ohms 

Letters  therefore  represent  numbers,  and  for 
any  letter  we  can  read  "any  number.”  When 
the  same  letter  occurs  again  in  the  same  ex- 
pression we  would  mentally  read  "the  same 
number,”  and  for  another  letter  "another  num- 
ber of  any  value.” 

These  letters  are  connected  by  the  usual  op- 
erational symbols  of  arithmetic,  -f,  — , X,  -h, 
and  so  forth.  In  algebra,  the  sign  for  division 
is  seldom  used,  a division  being  usually  written 
as  a fraction.  The  multiplication  sign,  X,  is 
usually  omitted  or  one  may  write  a period 
only.  Examples: 

2 X a X b = 2ob 
2.3.4.5a  = 2X3X4XSXa 

In  practical  applications  of  algebra,  an  ex- 
pression usually  states  some  physical  law  and 
each  letter  represents  a variable  quantity  which 
is  therefore  called  a variable.  A fixed  number 
in  front  of  such  a quantity  (by  which  it  is  to 
be  multiplied)  is  known  as  the  coefficient. 
Sometimes  the  coefficient  may  be  unknown,  yet 
to  be  determined;  it  is  then  also  written  as  a 
letter;  k is  most  commonly  used  for  this  pur- 
pose. 

The  Negative  In  ordinary  arithmetic  we 

Sign  seldom  work  with  negative 

numbers,  although  we  may 
be  "short”  in  a subtraction.  In  algebra,  how- 
ever, a number  may  be  either  negative  or  pos- 
itive. Such  a thing  may  seem  academic  but  a 
negative  quantity  can  have  a real  existence. 
We  need  only  refer  to  a debt  being  considered 
a negative  possession.  In  electrical  work,  how- 
ever, a result  of  a problem  might  be  a negative 
number  of  amperes  or  volts,  indicating  that  the 
direction  of  the  current  is  opposite  to  the  di- 
rection chosen  as  positive.  We  shall  have  il- 
lustrations of  this  shortly. 

Having  established  the  existence  of  negative 
quantities,  we  must  now  learn  how  to  work 
with  these  negative  quantities  in  addition,  sub- 
traction, multiplication  and  so  forth. 

In  addition,  a negative  number  Sdded  to  a 
positive  number  is  the  same  as  subtracting  a 
positive  number  from  it. 

7 7 

^ (add)  **  ~ (subtract) 

4 4 

or  we  might  write  it 


Similarly,  we  have: 

0 + ( — b)  = a — b 

When  a minus  sign  is  in  front  of  an  expres- 
sion in  brackets,  this  minus  sign  has  the  effect 
of  reversing  the  signs  of  every  term  within  the 
brackets: 

— (a  — b)  = — a + b 
— (2o  + 3b  - 5e)  = - 2a  — 3b  + 5e 

Multiplication.  When  both  the  multiplicand 
and  the  multiplier  are  negative,  the  product  is 
positive.  When  only  one  (either  one)  is  nega- 
tive the  product  is  negative.  The  four  possible 
cases  are  illustrated  below: 

+x+=+  +x-=- 

-X  + = - -X-  = -f- 

Division.  Since  division  is  but  the  reverse  of 
multiplication,  similar  rules  apply  for  the  sign 
of  the  quotient.  When  both  the  dividend  and 
the  divisor  have  the  same  sign  (both  negative 
or  both  positive)  the  quotient  is  positive.  If 
they  have  unlike  signs  (one  positive  and  one 
negative)  the  quotient  is  negative. 

+ _ + _ 

+ ~ - ~ 


Powers.  Even  powers  of  negative  numbers 
are  positive  and  odd  powers  are  negative.  Pow- 
ers of  positive  numbers  are  always  positive. 
Examples: 

-2»=-2X-2=  + 4 
-y=-2X-2X-2=+4X 
- 2 = - 8 

Roots.  Since  the  square  of  a negative  num- 
ber is  positive  and  the  square  of  a positive 
number  is  also  positive,  it  follows  that  a posi- 
tive number  has  two  square  roots.  The  square 
root  of  4 can  be  either  -f-2  or  —2  for  ( + 2) 
X ( + 2)  = +4  and  1-2)  X (-2)  = +4. 

Addition  and  Polynomials  are  quantities 

Subtraction  like  3ab’  -f  4ab''  — 7a’b’ 

which  have  several  terms  of 
different  names.  When  adding  polynomials, 
only  terms  of  the  same  name  can  be  taken  to- 
gether. 

7a=  + 8 ab"  + 3 a^b  -f  3 

a“  - 5 ab“  - b’ 


7 + (-3)  = 7-  3 = 4 


8a‘  -p  3 ab‘  -I-  3 a^b  - b'  -P  3 


HANDBOOK 


Division  751 


Collecting  terms.  When  an  expression  con- 
tains more  than  one  term  of  the  same  name, 
these  can  be  added  together  and  the  expression 
made  simpler: 

5 -|-  2 xy  + 3 xy'“  — 3 x’  + 7 xy  = 

5 x^  — 3 x’  -f-  2 xy  + 7 xy  + 3 xy’= 

2 x"  -f  9 xy  -f  3 xy= 

Multiplication  Multiplication  of  single  terms 
is  indicated  simply  by  writing 

them  together. 

a X b is  written  as  ob 

a X b’  is  written  as  ab’ 

Bracketed  quantities  are  multiplied  by  a 
single  term  by  multiplying  each  term: 

a (b  -f-  c + d)  = ab  ac  ad 

When  two  bracketed  quantities  ate  multi- 
plied, each  term  of  the  first  bracketed  quantity 
is  to  be  multiplied  by  each  term  of  the  second 
bracketed  quantity,  thereby  making  every  pos- 
sible combination. 

(o-f-b)  (c*4“d)  sc  ac  + od  + be  -4“  bd 

In  this  work  particular  care  must  be  taken 
to  get  the  signs  correct.  Examples: 

(a  + b)  (a  — b)  = a‘  + ab  — ab  — b’  = 
a’  - b^ 

(a  -1-  b)  (a  -b  b)  = a^  ab  -1-  ob  -1-  b'  = 
0=  + 2ab  -b  b^ 

(a  — b)  (a  — b)  = a'  — ab  — ob  -b  b’  = 
0=  - 2 ob  -b  b“ 

Division  It  is  possible  to  do  longhand  divi- 
sion in  algebra,  although  it  is 
somewhat  more  complicated  than  in  arithme- 
tic. However,  the  division  will  seldom  come 
out  even,  and  is  not  often  done  in  this  form. 
The  method  is  as  follows:  Write  the  terms  of 
the  dividend  in  the  order  of  descending  powers 
of  one  variable  and  do  likewise  with  the  di- 
visor. Example: 

Divide  5e“b  -b  21b“  -b  2o’  — 26ab"  by 
2a  - 3b 

Write  the  dividend  in  the  order  of  descending 
powers  of  a and  divide  in  the  same  way  as  in 
arithmetic. 


+ 4 ab  - 7b^ 

2o  - 3b  ) 2a*  + 5a^b  --  26  ab"  + 21b'" 
2a^  - 3a‘b 


-b  8a’b  - 26ab^ 
+ 8a’b  - 12ab= 


- 14ab=  -b  21b'' 

- 14ab''  + 21b'' 


Another  example:  Divide  x’  — y"  by  x — y 

X - y ) x’  -b  0 -b  0 - y"  ( x^  -b  xy  + / 
x“  — x-y 


+ - y= 

xy'  — y^ 


Factoring  Very  often  it  is  necessary  to  sim- 
plify expressions  by  finding  a fac- 
tor. This  is  done  by  collecting  two  or  more 
terms  having  the  same  factor  and  bringing  the 
factor  outside  the  brackets: 

6ab  -b  3ac  = 3o(2b  + e) 

In  a four  term  expression  one  can  take  to- 
gether two  terms  at  a time;  the  intention  is  to 
try  getting  the  terms  within  the  brackets  the 
same  after  the  factor  has  been  removed: 

30ac  - 18bc  -b  lOad  - 6bd  = 

6c  (5a  - 3b)  4-  2d  (5a  - 3b)  = 

(5a  - 3b)  (6c  -b  2d) 

Of  course,  this  is  not  always  possible  and 
the  expression  may  not  have  any  factors.  A 
similar  process  can  of  course  be  followed  when 
the  expression  has  six  or  eight  or  any  even 
number  of  terms. 

A special  case  is  a three-term  polynomial, 
which  can  sometimes  be  factored  by  writing 
the  middle  term  as  the  sum  of  two  terms: 

x=  - 7xy  -b  12/  may  be  rewritten  as 

x"  — 3xy  — 4xy  -b  1 2y^  = 

X (x  — 3y)  — 4y  (x  — 3y)  = 

(x  - 4y)  (x  - 3y) 

The  middle  term  should  be  split  into  two 
in  such  a way  that  the  sum  of  the  two  new 
terms  equals  the  original  middle  term  and  that 
their  product  equals  the  product  of  the  two 
outer  terms.  In  the  above  example  these  condi- 
tions are  fulfilled  for  — 3xy  — 4 xy  = — 7xy 
and  (-3xy)  (-4xy)  =;  12  xY.  It  is  not  al- 
ways possible  to  do  this  and  there  are  then  no 
simple  factors. 


752  Rad  io  Mathematics  and  Calculations  THE  RADIO 


Working  with 

Powers 
and  Roots 


When  two  powers  of  the 
same  number  are  to  be  mul- 
tiplied, the  exponents  are 
added. 


o‘  X o^  = oo  X ooo  = oooaa  = a‘  or 

a‘  X o’  = o'’*"'  = 0“ 


Roots  may  be  written  as  fractional  powers. 
Thus  may  be  written  as  a''^  because 

X = o 

and,  o'^  X = o’  = o 

Any  root  may  be  written  in  this  form 


b’  X b = b* 


\/T=b’^  ^=b’^  A/b’  = b’'* 


e’  X e'  = e“ 

Similarly,  dividing  of  powers  is  done  by 
subtracting  the  exponents. 


.=  a or-?:-=  a'’-’>  = o'  =« 


b" 

b» 


bbbbb  ,2  b‘ 
n.b-b-  = b^' 


rb'*-’*  = b’ 


The  same  notation  is  also  extended  in  the 
negative  direction: 


b-^  = — = 

" b’/i  — 


1 

Vb 


c-^ 


J 1_ 


Following  the  previous  rules  that  exponents 
add  when  powers  are  multiplied. 


Now  we  are  logically  led  into  some  impor- 
tant new  ways  of  notation.  We  have  seen  that 
when  dividing,  the  exponents  are  subtracted. 
This  can  be  continued  into  negative  exponents. 
In  the  following  series,  we  successively  divide 
by  a and  since  this  can  now  be  done  in  two 
ways,  the  two  ways  of  notation  must  have  the 
same  meaning  and  be  identical. 


X 

but  also  X 
therefore 

Powers  of  powers.  When  a power  is  again 
raised  to  a power,  the  exponents  are  multi- 
plied; 

(o*)’  = o"  (b-’)’  = b-’ 

(o’)*  = o“  (b-’)-’=b* 

This  same  rule  also  applies  to  roots  of  toots 
and  also  powers  of  roots  and  roots  of  powers 
because  a root  can  always  be  written  as  a frac- 
tional power. 


These  examples  illustrate  two  rules:  (1)  any 
number  raised  to  "zero’  power  equals  one  or 
unity;  (2)  any  quantity  raised  to  a negative 
power  is  the  inverse  or  reciprocal  of  the  same 
quantity  raised  to  the  same  positive  power. 


Roots.  The  product  of  the  square  root  of 
two  quantities  equals  the  square  root  of  their 
product. 

'/H  X Vb  = '/  ab 


v'^=-^Tfor 

Removing  radicals.  A root  or  radical  in  the 
denominator  of  a fraction  makes  the  expres- 
sion difficult  to  handle.  If  there  must  be  a rad- 
ical it  should  be  located  in  the  numerator 
rather  than  in  the  denominator.  The  removal 
of  the  radical  from  the  denominator  is  done 
by  multiplying  both  numerator  and  denomina- 
tor by  a quantity  which  will  remove  the  radi- 
cal from  the  denominator,  thus  rationalizing  it: 


Also,  the  quotient  of  two  roots  is  equal  to  the 
root  of  the  quotient. 


Note,  however,  that  in  addition  or  subtrac- 
tion the  square  root  of  the  sum  or  difference  is 
not  the  same  as  the  sum  or  difference  of  the 
square  roots. 

Thus,  vT-  vT  = 3-2=1 
but  \/9  - 4 = VT=  2.2361 
Likewise  -(-  Vb  is  not  the  same  as  a -f  b 


-J ±3 1 f- 

'/n  yfi  X \/a  a ^ a 

Suppose  we  have  to  rationalize 
3a 

In  this  case  we  must  multiply 

numerator  and  denominator  by  V~i  — the 
same  terms  but  with  the  second  having  the 
opposite  sign,  so  that  their  product  will  not 
contain  a root. 

3a  3a(  v’g  — \/b)  3a(\/o  — y'^ 

\/a -f  ( vV-|-\/b)(\/a  — x/bl  o — b 


HANDBOOK 


Powers,  Roots,  Imoginaries  753 


Imaginary  Since  the  square  of  a negative 
Numbers  number  is  positive  and  the 

square  of  a positive  number  is 
also  positive,  the  square  root  of  a negative 
number  can  be  neither  positive  not  negative. 
Such  a number  is  said  to  be  imaginary;  the 
most  common  such  number  ( V — 1 ) is  often 
represented  by  the  letter  i in  mathematical 
work  or  j in  electrical  work. 


be  accounted  for  separately,  has  found  a sym- 
bolic application  in  vector  notation.  These  are 
coveted  later  in  this  chapter. 

Equations  of  the  Algebraic  expressions  usu- 
First  Degree  ally  come  in  the  form  of 

equations,  that  is,  one  set 
of  terms  equals  another  set  of  terms.  The  sim- 
plest example  of  this  is  Ohm’s  Law: 


\/  — 1 zz  i or  j and  i*  or  i‘  zz  — 1 

Imaginary  numbers  do  not  exactly  corre- 
spond to  anything  in  our  experience  and  it  is 
best  not  to  try  to  visualize  them.  Despite  this 
fact,  their  interest  is  much  more  than  academ- 
ic, for  they  are  extremely  useful  in  many  cal- 
culations involving  alternating  currents. 

The  square  root  of  any  other  negative  num- 
ber may  be  reduced  to  a product  of  two  roots, 
one  positive  and  one  negative.  For  instance: 

V^^57  = = ivTr 

or,  in  general 

V -a  = i \fa 


E = IR 

One  of  the  three  quantities  may  be  unknown 
but  if  the  other  two  are  known,  the  third  can 
be  found  readily  by  substituting  the  known 
values  in  the  equation.  This  is  very  easy  if  it 
is  E in  the  above  example  that  is  to  be  found; 
but  suppose  we  wish  to  find  / while  E and  R 
are  given.  We  must  then  rearrange  the  equa- 
tion so  that  / comes  to  stand  alone  to  the  left 
of  the  equality  sign.  This  is  known  as  solving 
the  equation  for  L 

Solution  of  the  equation  in  this  case  is  done 
simply  by  transposing.  If  two  things  are  equal 
then  they  must  still  be  equal  if  both  are  multi- 
lied  or  divided  by  the  same  number.  Dividing 
oth  sides  of  the  equation  by  R: 


Since  i = V — 1,  the  powers  of  i have  the 
following  values: 

F = -1 

F = -1  X i = -i 

F = +1 

i“  = + I X i = i 

Imaginary  numbers  are  different  from  either 
positive  or  negative  numbers;  so  in  addition  or 
subtraction  they  must  always  be  accounted  for 
separately.  Numbers  which  consist  of  both  real 
and  imaginary  parts  are  called  complex  num- 
bers. Examples  of  complex  numbers: 


If  it  were  required  to  solve  the  equation  for 
R,  we  should  divide  both  sides  of  the  equation 
by  7, 

-j-  = R or  R = -|- 

A little  more  complicated  example  is  the 
equation  for  the  reactance  of  a condenser: 

X ! 

To  solve  this  equation  for  C,  we  may  multi- 
ply both  sides  of  the  equation  by  C and  divide 
both  sides  by  X 


3 -f  4i  = 3 -f  4 
a -4-  bi  zz  a + b V — 1 


X 


! X 

2Tif  C 
1 

2iif  X 


C 

X' 


or 


Since  an  imaginary  number  can  never  be 
equal  to  a real  number,  it  follows  that  in  an 
equality  like 

a + bi  = e + di 

a must  equal  e and  bi  must  equal  di 

Complex  numbers  are  handled  in  algebra 
just  like  any  other  expression,  considering  i 
as  a known  quantity.  Whenever  powers  of  i 
occur,  they  can  be  replaced  by  the  equivalents 
given  above.  This  idea  of  having  In  one  equa- 
tion two  separate  sets  of  quantities  which  must 


This  equation  is  one  of  those  which  requires 
a good  knowledge  of  the  placing  of  the  deci- 
mal point  when  solving.  Therefore  we  give  a 
few  examples:  What  is  the  reactance  of  a 25 
HniA.  capacitor  at  1000  kc..’  In  filling  in  the 
given  values  in  the  equation  we  must  remem- 
ber that  the  units  usea  are  farads,  cycles,  and 
ohms.  Hence,  we  must  write  25  /x^ifd.  as  25 
millionths  of  a millionth  of  a farad  or  25  x 
10  “ farad;  similarly,  1000  kc.  must  be  con- 
verted to  1,000,000  cycles.  Substituting  these 
values  in  the  original  equation,  we  have 


754  Radio  Mathematics  and  Calculations  THE  RADIO 


2 X 3.14  X 1,000,000  X 25  xT»-» 

y 1 10« 

6.28  X 10«  X 2S  X 10-“  ““  6.28x25 
= 6360  ohms 

A bias  resistor  of  1000  ohms  should  be  by- 
passed, so  that  at  the  lowest  frequency  the  re- 
actance of  the  condenser  is  1/1 0th  of  that  of 
the  resistor.  Assume  the  lowest  frequency  to  be 
50  cycles,  then  tbe  required  capacity  should 
have  a reactance  of  100  ohms,  at  50  cycles; 


It  is,  however,  simpler  in  this  case  to  use 
the  elimination  method  Multiply  both  sides  of 
the  first  equation  by  two  and  add  it  to  the 
second  equation: 

6x  + lOy  = 14 
4x  - lOy  =z  3 

add 

lOx  =17  X = 1.7 

Substituting  this  value  of  x in  the  first  equa- 
tion, we  have 


2x3.14x50x  100 

<=  = 6:28^1000  '"'‘'“'“'■“‘I* 

C = 32  mW- 

In  the  third  possible  case,  it  may  be  that  the 
frequency  is  the  unknown.  This  happens  for 
instance  in  some  tone  control  problems.  Sup- 
pose it  is  required  to  find  the  frequency  which 
makes  the  reactance  of  a 0.03  ;ifd.  condenser 
equal  to  100,000  ohms. 

First  we  must  solve  the  equation  for  /.  This 
is  done  by  transposition. 

Y _ ^ f — I 

2i«fC  ^ — 2TrCX 

Substituting  known  values 

* ~ 2 X 3.14  X 0.03  X 10-"  x 100,000  *7'’®* 

^ = -oloisax  '7*'®*  = 53  C7cles 

These  equations  are  known  as  first  degree 
equations  with  one  unknown.  First  degree,  be- 
cause the  unknown  occurs  only  as  a first  power. 
Such  an  equation  always  has  one  possible  so- 
lution or  roof  if  all  the  other  values  are  known. 

If  there  are  two  unknowns,  a single  equa- 
tion will  not  suffice,  for  there  are  then  an  infi- 
nite number  of  possible  solutions.  In  the  case 
of  two  unknowns  we  need  two  independent 
simultaneous  equations.  An  example  of  this  is: 

3x  5y  = 7 4x  — lOy  = 3 

Required,  to  find  x and  y. 

This  type  of  work  is  done  either  by  the  sub- 
stitution method  or  by  the  elimination  method. 
In  the  substitution  method  we  might  write  for 
the  first  equation; 

3x  = 7 - 5y  .-.  X = 

(The  symbol  .’.  menns,  therefore  or  hence). 

This  value  of  x can  then  be  substituted  for  x 
in  the  second  equation  making  it  a single  equa- 
tion with  but  one  unknown,  y. 


5.1  + 5y  = 7 .-.  5y  = 7 - 5.1  = 1.9  .-. 
y = 0.38 


Figure  3. 

In  this  simple  network  the  current  diV/dei 
through  the  2000-ohm  anti  3000-ohm  resistors. 
The  current  through  each  may  be  found  by 
using  two  simultaneous  linear  equations.  Note 
that  the  arrows  indicate  the  direction  of  elec- 
tron flow  os  explained  on  page  16. 


An  application  of  two  simultaneous  linear 
equations  will  now  be  given.  In  Figure  3 a 
simple  network  is  shown  consisting  of  three  re- 
sistances; let  it  be  required  to  find  the  currents 
Ii  and  h in  the  two  branches. 

The  general  way  in  which  all  such  prob- 
lems can  be  solved  is  to  assign  directions  to 
the  currents  through  the  various  resistances. 
When  these  are  chosen  wrong  it  will  do  no 
harm  for  the  result  of  the  equations  will  then 
be  negative,  showing  up  the  error.  In  this  sim- 
ple illustration  there  is,  of  course,  no  such  dif- 
ficulty. 

Next  we  write  the  equations  for  the  meshes, 
in  accordance  with  Kirchhoff’s  second  law.  All 
voltage  drops  in  the  direction  of  the  curved 
arrow  are  considered  positive,  the  reverse  ones 
negative.  Since  there  are  two  unknowns  we 
write  two  equations. 

1000  (li  + I,)  -F  2000  I,  = 6 
-2000  I,  -f  3000  1:  = 0 
Expand  the  first  equation 

3000  I,  -4-  1000  I,  = 6 


HANDBOOK 


Quadratic  Equations  755 


A MORE  complicated  PROB- 
LEM REQUIRING  THE  SOLUTION 
OF  CURRENTS  IN  A NETWORK. 


This  problem  is  simitar  to  that  in  Figaro  3 but 
requires  the  use  of  three  simuitaneous  linear 
equations. 


Multiply  this  equation  by  3 

9000  li  + 3000  I,  = 18 
Subtracting  the  second  equation  from  the  first 

11000  h = 18 

I,  = 18/11000  = 0.00164  amp. 

Filling  in  this  value  in  the  second  equation 

3000  h = 3.28  h = 0.00109  amp. 

A similar  problem  but  requiring  three  equa- 
tions is  shown  in  Figure  4,  This  consists  of  an 
unbalanced  bridge  and  the  problem  is  to  find 
the  current  in  the  bridge-branch,  k We  again 
assign  directions  to  the  different  currents, 
guessing  at  the  one  marked  L,  The  voltages 
around  closed  loops  ABC  [eq.  (1)]  and  BDC 
[eq.  (2)]  equal  zero  and  are  assumed  to  be 
positive  in  a counterclockwise  direction;  that 
from  D to  A equals  10  volts  [eq.  (3)]. 

(1 ) 

- 1000  I.  -1-  2000  h — 1000  L = 0 

(2) 

— 1000  (I,— Is) -1-1000  ls-l-3000  (Is-|-l3)=0 

(3) 

1000  I,  -1-  1000  (I,  - Is)  — 10  = 0 

Expand  equations  (2)  and  (3) 

(2) 

— 1000  h + 3000  Is  + 5000  Is  = 0 

(3) 

2000  h - 1000  Is  - 10  = 0 

Subtract  equation  (2)  from  equation  (1) 

(a) 

- 1 000  Is  - 6000  Is  = 0 


Multiply  the  second  equation  by  2 and  add 
it  to  the  third  equation 

(b) 

6000  It  -I-  9000  Is  - 10  = 0 

Now  we  have  but  two  equations  with  two 
unknowns. 

Multiplying  equation  (a)  by  6 and  adding 
to  equation  (b)  we  have 

-27000  Is  - 10  = 0 
I,  = -10/27000  = —0.00037  amp. 

Note  that  now  the  solution  is  negative 
which  means  that  we  have  drawn  the  arrow 
for  Is  in  Figure  4 in  the  wrong  direction.  The 
current  is  0.37  ma.  in  the  other  direction. 

Second  Degree  or  A somewhat  similar 
Quadratic  Equations  problem  in  radio  would 
be,  if  power  in  watts 
and  resistance  in  ohms  of  a circuit  are  given, 
to  find  the  voltage  and  the  current.  Example: 
When  lighted  to  normal  brilliancy,  a 100  watt 
lamp  has  a resistance  of  49  ohms;  for  what 
line  voltage  was  the  lamp  designed  and  what 
current  would  it  take. 

Here  we  have  to  use  the  simultaneous  equa- 
tions: 

P = El  and  E = IR 

Filling  in  the  known  values: 

P = El  = 100  and  E = IR  = I X 49 

Substitute  the  second  equation  into  the  first 
equation 

P = El  = (I)  X I X 49  = 49  P = 100 

.-.  I = a/M=-^  = 1.43  amp. 

Substituting  the  found  value  of  1.43  amp.  for 
/ in  the  first  equation,  we  obtain  the  value  of 
the  line  voltage,  70  volts. 

Note  that  this  is  a second  degree  equation 
for  we  finally  had  the  second  power  of  /.  Also, 
since  the  current  in  this  problem  could  only  be 
positive,  the  negative  square  root  of  100/49 
or  — 10/7  was  not  used.  Strictly  speaking, 
however,  there  are  two  more  values  that  sat- 
isfy both  equations,  these  ate  —1.43  and  —70. 

In  general,  a second  degree  equation  in  one 
unknown  has  two  roots,  a third  degree  equa- 
tion three  roots,  etc. 

The  Quadratic  Quadratic  or  second  degree 
Equation  equations  with  but  one  un- 

known can  be  reduced  to  the 

general  form 

ax"  -F  bx  -f  e = 0 


756  Radio  Mathematics  and  Calculations  THE  RADIO 


where  x is  the  unknown  and  a,  b,  and  c are 

constants. 

This  type  of  equation  can  sometimes  be 
solved  by  the  method  of  factoring  a three- 
term  expression  as  follows: 


2x‘  + 7x  + 6 = 0 


factoring: 


2:d  -H  4x  -f  3x  6 = 0 
2x  (x  + 2)  -h  3 (x  -I-  2)  = 0 


(2x  -(-3)  (x  -f  2)  = 0 

There  are  two  possibilities  when  a product 
is  aero.  Either  the  one  or  the  other  factor 
equals  zero.  Therefore  there  are  two  solutions. 

2x.  -h  3 = 0 X,  + 2 = 0 

2xi  = -3  X,  = -2 


X.  = -I'/a 

Since  factoring  is  not  always  easy,  the  fol- 
lowing general  solution  can  usually  be  em- 
ployed; in  this  equation  a,  h,  and  c are  the  co- 
efficients referred  to  above. 


Also:  (Xi,  - Xo)*  = r - R’ 
and  ± (Xl  Xo)  = V Z’  - R’ 

But  here  we  do  not  know  the  sign  of  the  so- 
lution unless  there  are  other  facts  which  indi- 
cate it.  To  find  either  Xl  or  Xo  alone  it  would 
have  to  be  known  whether  the  one  or  the  other 
is  the  larger. 


Logarifhms 

Definition  A logarithm  is  the  power  (or  ex- 
ond  Use  ponent)  to  which  we  must  raise 
one  number  to  obtain  another. 
Although  the  large  numbers  used  in  logarith- 
mic work  may  make  them  seem  difficult  or 
complicated,  in  reality  the  principal  use  of 
logarithms  is  to  simplify  calculations  which 
would  otherwise  be  extremely  laborious. 

We  have  seen  so  far  that  every  operation 
in  arithmetic  can  be  reversed.  If  we  have  the 
addition: 

0 b = c 

we  can  reverse  this  operation  in  two  ways.  It 
may  be  that  b is  the  unknown,  and  then  we 
reverse  the  equation  so  that  it  becomes 


V — b ± V b’  — 4a« 

A _ — 

Applying  this  method  of  solution  to  the  pre- 
vious example: 

V _-7±V49  - 8 X 6_  -r±va_-7±l 

A - _ ^ 

Xl  = ' = - 1 Vi 


A practical  example  involving  quadratics  is 
the  law  of  impedance  in  a.c.  circuits.  However, 
this  is  a simple  kind  of  quadratic  equation 
which  can  be  solved  readily  without  the  use 
of  the  special  formula  given  above. 

Z = VR’  -I-  (Xl  - Xo)* 

This  equation  can  always  be  solved  for  R, 
by  squaring  both  sides  of  the  equation.  It 
should  now  be  understood  that  squaring  both 
sides  of  an  equation  as  well  as  multiplying 
both  sides  with  a term  containing  the  unknown 
may  add  a new  root.  Since  we  know  here  that 
Z and  R are  positive,  when  we  square  the  ex- 
pression there  is  no  ambiguity. 

2’  = R*  -I-  (Xl  - Xo)' 
and  R'  = r - (Xl  - Xo)' 
orR  = V Z'  - (Xl  - Xo)‘ 


c — 0 = b 

It  is  also  possible  that  we  wish  to  know  a,  and 
that  b and  c are  given.  The  equation  then  be- 
comes 

c — b = 0 

We  call  both  of  these  reversed  operations  sub- 
traction, and  we  make  no  distinction  between 
the  two  possible  reverses. 

Multiplication  can  also  be  reversed  in  two 
manners.  In  the  multiplication 

eb  = e 

we  may  wish  to  know  a,  when  b and  c are 
given,  or  we  may  wish  to  know  b when  a and 
c are  given.  In  both  cases  we  speak  of  division, 
and  we  make  again  no  distinction  between  the 
two. 

In  the  case  of  powers  we  can  also  reverse 
the  operation  in  two  manners,  but  now  they 
are  not  equivalent.  Suppose  we  have  the  equa- 
tion 

a"  = c 

If  a is  the  unknown,  and  b and  c are  given, 
we  may  reverse  the  operation  by  writing 

^=0 

This  operation  we  call  taking  the  root.  But 
there  is  a third  possibility:  that  a and  c are 
given,  and  that  we  wish  to  know  b.  In  other 


HANDBOOK 


Logarithms  757 


words,  the  question  is  "to  which  power  must 
we  raise  a so  as  to  obtain  c?”.  This  operation 
is  known  as  taking  the  logarithm,  and  b is  the 
logarithm  of  c to  the  base  a.  We  write  this 
operation  as  follows: 

log.  c = b 

Consider  a numerical  example.  Wc  know  2*=8. 
We  can  reverse  this  operation  by  asking  "to 
which  power  must  we  raise  2 so  as  to  obtain 
8?”  Therefore,  the  logarithm  of  8 to  the  base 
2 is  3,  or 

log.  8 = 3 

Taking  any  single  base,  such  as  2,  we  might 
write  a series  of  all  the  powers  of  the  base  next 
to  the  series  of  their  logarithms: 

Number;  2 4 8 16  32  64  128  256  512  1024 
Logarithm:  123456  7 8 9 10 

We  can  expand  this  table  by  finding  terms 
between  the  terms  listed  above.  For  instance, 
if  we  let  the  logarithms  increase  with  1/2  each 
time,  successive  terms  in  the  upper  series 
would  have  to  be  multiplied  by  the  square  root 
of  2.  Similarly,  if  we  wish  to  increase  the  log- 
arithm by  1/10  at  each  term,  the  ratio  between 
two  consecutive  terms  in  the  upper  series 
would  be  the  tenth  root  of  2.  Now  this  short 
list  of  numbers  constitutes  a small  logarithm 
table.  It  should  be  clear  that  one  could  find 
the  logarithm  of  any  number  to  the  base  2. 
This  logarithm  will  usually  be  a number  with 
many  decimals. 

Logarithmic  The  fact  that  we  chose  2 as 
Boses  a base  for  the  illustration  is 

purely  arbitrary.  Any  base 
could  be  used,  and  therefore  there  are  many 
possible  systems  of  logarithms.  In  practice  we 
use  only  two  bases:  The  most  frequently  used 
base  is  10,  and  the  system  using  this  base  is 
known  as  the  system  of  common  logarithms, 
or  Briggs’  logarithms.  The  second  system  em- 
ploys as  a base  an  odd  number,  designated  by 
the  letter  e;  e = 2.71828.  . . . This  is  known 
as  the  natural  logarithmic  system,  also  as  the 
Napierian  system,  and  the  hyperbolic  system. 
Although  different  writers  may  vary  on  the 
subject,  the  usual  notation  is  simply  log  a for 
the  common  logarithm  of  a,  and  log^  a (or 
sometimes  In  a)  for  the  natural  logarithm  of 
a.  We  shall  use  the  common  logarithmic  sys- 
tem in  most  cases,  and  therefore  we  shall  ex- 
amine this  system  more  closely. 

Common  In  the  system  wherein  10  is  the 
Logarithms  base,  the  logarithm  of  10  equals 
1;  the  logarithm  of  100  equals  2, 
etc.,  as  shown  in  the  following  table; 


log  10  = log  10'  = 1 

log  100  = log  10"  = 2 

log  1,000  = log  10"  = 3 

log  10,000  = log  10*  = 4 

log  100,000  = log  10°  = 5 

log  1,000,000  = log  10°  = 6 

This  table  can  be  extended  for  numbers  less 
than  10  when  we  remember  the  rules  of  pow- 
ers discussed  under  the  subject  of  algebra. 
Numbers  less  than  unity,  too,  can  be  written 
as  powers  of  ten. 


log  1 

= log 

10”  = 

0 

log  0.1 

= log 

10’=  ■ 

-1 

log  0.01 

= log 

10°  = ■ 

-2 

log  0.001 

= log 

10°  = 

-3 

log  0.0001 

= log 

10  - = ■ 

-4 

From  these  examples  follow  several  rules: 
The  logarithm  of  any  number  between  aero 
and  + 1 is  negative;  the  logarithm  of  zero  is 
minus  infinity;  the  logarithm  of  a number 
greater  than  -f  1 is  positive.  Negative  num- 
bers have  no  logarithm.  These  rules  are  true 
of  common  logarithms  and  of  logarithms  to 
any  base. 

The  logarithm  of  a number  between  the 
powers  of  ten  is  an  irrational  number,  that  is, 
it  has  a never  ending  series  of  decimals.  For  in- 
stance, the  logarithm  of  20  must  be  between 
1 and  2 because  20  is  between  10  and  100;  the 
value  of  the  logarithm  of  20  is  1.30103.  . . . 
The  part  of  the  logarithm  to  the  left  of  the 
decimal  point  is  called  the  characteristic,  while 
the  decimals  are  called  the  mantissa.  In  the 
case  of  1.30103  . .,  the  logarithm  of  20,  the 
characteristic  is  1 and  the  mantissa  is  .30103  . . 

Properties  of  If  the  base  of  our  system  is 
Logarithms  ten,  then,  by  definition  of  a 
logarithm: 

lO'or  • = a 

or,  if  the  base  is  raised  to  the  power  having  an 
exponent  equal  to  the  logarithm  of  a number, 
the  result  is  that  number. 

The  logarithm  of  a product  is  equal  to  the 
sum  of  the  logarithms  of  the  two  factors. 

log  ob  = log  a + log  b 

This  is  easily  proved  to  be  true  because,  it 


Figure  5.  FOUR  PLACE  LOGARITHM  TABLES. 


N 

0 

1 

2 

3 

4 

5 

6 

1 ^ 

8 1 

9 

N 

1 ^ 

I 1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

0000 

0043 

0086 

0128 

0170 

0212 

0253 

0294 

0334 

0374 

55 

7404 

7412 

7419 

7427 

7435 

7443 

7451 

7459 

7466 

7474 

11 

0414 

0453 

0492 

0531 

0569 

0607 

0645 

0682 

0719 

0755 

56 

7482 

7490 

7497 

7505 

7513 

7520 

7528 

7536 

7543 

7551 

12 

0792 

0828 

0864 

0899 

0934 

0969 

1004 

1038 

1072 

1106 

57 

7559 

7566 

7574 

7582 

7589 

7597 

7604 

7612 

7619 

7527 

13 

1139 

1173 

1206 

1239 

1271 

1303 

1335 

1367 

1399 

1430 

58 

7634 

7642 

7649 

7657 

7664 

7672 

7679 

7686 

7694 

7701 

14 

1461 

1492 

1523 

1553 

1584 

1614 

1644 

1673 

1703 

1732 

59 

7709 

7716 

7723 

7731 

7738 

7745 

7752 

7760 

7767 

7774 

15 

1761 

1790 

1818 

1847 

1875 

1903 

1931 

1959 

1987 

2014 

60 

7782 

7789 

7796 

7803 

7810 

7818 

7825 

7832 

7839 

7846 

16 

2041 

2068 

2095 

2122 

2148 

2175 

2201 

2227 

2253 

2279 

61 

7853 

7860 

7868 

7875 

7882 

7889 

7896 

7903 

7910 

7917 

17 

2304 

2330 

2355 

2380 

2405 

2430 

2455 

2480 

2504 

2529 

62 

7924 

7931 

7938 

7945 

7952 

7959 

7966 

7973 

7980 

7987 

18 

2553 

2577 

2601 

2625 

2648 

2672 

2695 

2718 

2742 

2765 

63 

7993 

8000 

8007 

8014 

8021 

8028 

8035 

8041 

8048 

8055 

19 

2788 

2810 

2833 

2856 

2878 

2900 

2923 

2945 

2967 

2989 

64 

8062 

8069 

8075 

8082 

8089 

8096 

8102 

8109 

8116 

8122 

20 

3010 

3032 

3054 

3075 

3096 

3118 

3139 

3160 

3181 

3201 

65 

8129 

8136 

8142 

8149 

8156 

8162 

8169 

8176 

8182 

8189 

21 

3222 

3243 

3263 

3284 

3304 

3324 

3345 

3365 

3385 

3404 

66 

8195 

8202 

8209 

8215 

8222 

8228 

8235 

8241 

8248 

8254 

22 

3424 

3444 

3464 

3483 

3502 

3522 

3541 

3560 

3579 

3598 

67 

8261 

8267 

8274 

8280 

8287 

8293 

8299 

8306 

8312 

8319 

23 

3617 

3636 

3655 

3674 

3692 

3711 

3729 

3747 

3766 

3784 

68 

8325 

8331 

8338 

8344 

8351 

8357 

8363 

8370 

8376 

8382 

24 

3802 

3820 

3838 

3856 

3874 

3892 

3909 

3927 

3945 

3962 

69 

8388 

8395 

8401 

8407 

8414 

8420 

8426 

8432 

8439 

8445 

25 

3979 

3997 

4014 

4031 

4048 

4065 

4082 

4099 

4116 

4133 

70 

8451 

8457 

8463 

8470 

8476 

8482 

8488 

8494 

8500 

8506 

26 

4150 

4166 

4183 

4200 

4216 

4232 

4249 

4265 

4281 

4298 

71 

8513 

8519 

8525 

8531 

8537 

8543 

8549 

8555 

8561 

8567 

27 

4314 

4330 

4346 

4362 

4378 

4393 

4409 

4425 

4440 

4456 

72 

8573 

8579 

8585 

8591 

8597 

8603 

8609 

8615 

8621 

8627 

28 

4472 

4487 

4502 

4518 

4533 

4548 

4564 

4579 

4594 

4609 

73 

8633 

8639 

8645 

8651 

8657 

8663 

8669 

8675 

8681 

8686 

29 

4624 

4639 

4654 

4669 

4683 

4698 

4713 

4728 

4742 

4757 

74 

8692 

8698 

8704 

8710 

8716 

8722 

8727 

8733 

8739 

8745 

30 

4771 

4786 

4800 

4814 

4829 

4843 

4857 

4871 

4886 

4900 

75 

8751 

8756 

8762 

8768 

8774 

8779 

8785 

8791 

8797 

8802 

31 

4914 

4928 

4942 

4955 

4969 

4983 

4997 

5011 

5024 

5038 

76 

8808 

8814 

8820 

8825 

8831 

8837 

8842 

8848 

8854 

8859 

32 

5051 

5065 

5079 

5092 

5105 

5119 

5132 

5145 

5159 

5172 

77 

8865 

8871 

8876 

8882 

8887 

8893 

8899 

8904 

8910 

8915 

33 

5185 

5198 

5211 

5224 

5237 

5250 

5263 

5276 

5289 

5302 

78 

8921 

8927 

8932 

8938 

8943 

8949 

8954 

8960 

8965 

8971 

34 

5315 

5328 

5340 

5353 

5366 

5378 

5391 

5403 

5416 

5428 

79 

8976 

8982 

8987 

8993 

8998 

9004 

9009 

9015 

9020 

9025 

35 

5441 

5453 

5465 

5478 

5490 

5502 

5514 

5527 

5539 

5551 

80 

9031 

9036 

9042 

9047 

9053 

9058 

9063 

9069 

9074 

9079 

36 

5563 

5575 

5587 

5599 

5611 

5623 

5635 

5647 

5658 

5670 

81 

9085 

9090 

9096 

9101 

9106 

9112 

9117 

9122 

9128 

9133 

37 

5682 

5694 

5705 

5717 

5729 

5740 

5752 

5763 

5775 

5786 

82 

9138 

9143 

9149 

9154 

9159 

9165 

9170 

9175 

9180 

9186 

38 

5798 

5809 

5821 

5832 

5843 

5855 

5866 

5877 

5888 

5899 

83 

9191 

9196 

9201 

9206 

9212 

9217 

9222 

9227 

9232 

9238 

39 

5911 

5922 

5933 

5944 

5955 

5966 

5977 

5988 

5999 

6010 

84 

9243 

9248 

9253 

9258 

9263 

9269 

9274 

9279 

9284 

9289 

40 

6021 

6031 

6042 

6053 

6064 

6075 

6085 

6096 

6107 

6117 

85 

9294 

9299 

9304 

9309 

9315 

9320 

9325 

9330 

9335 

9340 

41 

6128 

6138 

6149 

6160 

6170 

6180 

6191 

6201 

6212 

6222 

86 

9345 

9350 

9355 

9360 

9365 

9370 

9375 

9380 

9385 

9390 

42 

6232 

6243 

6253 

6263 

6274 

6284 

6294 

6304 

6314 

6325 

87 

9395 

9400 

9405 

9410 

9415 

9420 

9425 

9430 

9435 

9440 

43 

6335 

6345 

6355 

6365 

6375 

6385 

6395 

6405 

6415 

6425 

88 

9445 

9450 

9455 

9460 

9465 

9469 

9474 

9479 

9484 

9489 

44 

6435 

6444 

6454 

6464 

6474 

6484 

6493 

6503 

6513 

6522 

89 

9494 

9499 

9504 

9509 

9513 

9518 

9523 

9528 

9533 

9538 

45 

6532 

6542 

6551 

6561 

6571 

6580 

6590 

6599 

6609 

6618 

90 

9542 

9547 

9552 

9557 

9562 

9566 

9571 

9576 

9581 

9586 

46 

6628 

6637 

6646 

6656 

6665 

6675 

6684 

6693 

6702 

6712 

91 

9590 

9595 

9600 

9605 

9609 

9614 

9619 

9624 

9628 

9633 

47 

6721 

6730 

6739 

6749 

6758 

6767 

6776 

6785 

6794 

6803 

92 

9638 

9643 

9647 

9652 

9657 

9661 

9666 

9671 

9675 

9680 

48 

6812 

6821 

6830 

6839 

6848 

6857 

6866 

6875 

6884 

6893 

93 

9685 

9689 

9694 

9699 

9703 

9708 

9713 

9717 

9722 

9727 

49 

6902 

6911 

6920 

6928 

6937 

6946 

6955 

6964 

6972 

6981 

94 

9731 

9736 

9741 

9745 

9750 

9754 

9759 

9763 

9768 

9773 

50 

6990 

6998 

7007 

7016 

7024 

7033 

7042 

7050 

7059 

7067 

95 

9777 

9782 

9786 

9791 

9795 

9800 

9805 

9809 

9814 

9818 

51 

7076 

7084 

7093 

7101 

7110 

7118 

7126 

7135 

7143 

7152 

96 

9823 

9827 

9832 

9836 

9841 

9845 

9850 

9854 

9859 

9863 

52 

7160 

7168 

7177 

7185 

7193 

7202 

7210 

7218 

7226 

7235 

97 

9868 

9872 

9877 

9881 

9886 

9890 

9894 

9899 

9903 

9908 

53 

7243 

7251 

7259 

7267 

7275 

7284 

7292 

7300 

7308 

7316 

98 

9912 

9917 

9921 

9926 

9930 

9934 

9939 

9943 

9948 

9952 

54 

7324 

7332 

7340 

7348 

7356 

7364 

7372 

7380 

7388 

7396 

99 

9956 

9961 

9965 

9969 

9974 

9978 

9983 

9987 

9991 

9996 

H 

X 

m 

> 

U 

o 


758  Radio  Mathematics  and  Calculations 


HANDBOOK 


Logarithm  Tables  759 


was  shown  before  that  when  multiplying  to 
powers,  the  exponents  are  added;  therefore, 

q ^ ^ ^ ^ X ^ **“10 

Similarly,  the  logarithm  of  a quotient  is  the 
difference  between  the  logarithm  of  the  divi- 
dend and  the  logarithm  of  the  divisor. 

log-^  = log  a — log  b 

This  is  so  because  by  the  same  rules  of  ex- 
ponents: 

_? ^ I Q(  loK  ft  - log  b) 

b — lO'bEi  — '' 

We  have  thus  established  an  easier  way  of 
multiplication  and  division  since  these  opera- 
tions have  been  reduced  to  adding  and  sub- 
tracting. 

The  logarithm  of  a power  of  a number  is 
equal  to  the  logarithm  of  that  number,  multi- 
plied by  the  exponent  of  the  power. 

log  O'  = 2 log  a and  log  o^  ==  3 log  o 

or,  in  general; 

log  o"  = n log  o 

Also,  the  logarithm  of  a root  of  a number 
is  equal  to  the  logarithm  of  that  number  di- 
vided by  the  index  of  the  root: 

log  V~a  =-L|og  o 

It  follows  from  the  rules  of  multiplication, 
that  numbers  having  the  same  digits  but  dif- 
ferent locations  for  the  decimal  point,  have 
logarithms  with  the  same  mantissa: 

log  829  = 2.918555 

log  82.9  = 1.918555 

log  8.29  = 0.918555 

log  0.829  = -1.918555 

log  0.0829  = -2.918555 

log  829  = log  (8.29  X 100)  = log  8.29  + 
log  100  = 0.918555  + 2 

Logarithm  tables  give  the  mantissas  of  log- 
arithms only.  The  characteristic  has  to  be  de- 
termined by  inspection.  The  characteristic  is 
equal  to  the  number  of  digits  to  the  left  of  the 
decimal  point  minus  one.  In  the  case  of  loga- 
rithms of  numbers  less  than  unity,  the  charac- 
teristic is  negative  and  is  equal  to  the  number 
of  ciphers  to  the  right  of  the  decimal  point 
plus  one. 

For  reasons  of  convenience  in  making  up 


logarithm  tables,  it  has  become  the  rule  that 
the  mantissa  should  always  be  positive.  Such 
notations  above  as  —1.918555  really  mean 
( + 0.918555  -1)  and  -2.981555  means 

(+  0.918555  — 2).  There  are  also  some  other 
notations  in  use  such  as 

T.918555  and  2.918555 

also  9.918555  - 10  8.918555  -10 

7.918555  - 10,  etc. 

When,  after  some  addition  and  subtraction 
of  logarithms  a mantissa  should  come  out  neg- 
ative, one  cannot  look  up  its  equivalent  num- 
ber or  anti-logarithm  in  the  table.  The  man- 
tissa must  first  be  made  positive  by  adding  and 
subtracting  an  appropriate  integral  number. 
Example:  Suppose  we  find  that  the  logarithm 
of  a number  is  —0.34569,  then  we  can  trans- 
form it  into  the  proper  form  by  adding  and 
subtracting  1 

1 -1 
-0.34569 

0.65431  -1  or  -1.65431 

Using  Logarithm  Tobies 

Logarithms  are  used  for  calculations  involv- 
ing multiplication,  division,  powers,  and  roots. 
Especially  when  the  numbers  are  large  and  for 
higher,  or  fractional  powers  and  roots,  this  be- 
comes the  most  convenient  way. 

Logarithm  tables  are  available  giving  the 
logarithms  to  three  places,  some  to  four  places, 
others  to  five  and  six  places.  The  table  to  use 
depends  on  the  accuracy  required  in  the  result 
of  our  calculations.  The  four  place  table, 
printed  in  this  chapter,  permits  the  finding  of 
answers  to  problems  to  four  significant  figures 
which  is  good  enough  for  most  constructional 
purposes.  If  greater  accuracy  is  required  a five 
place  table  should  be  consulted.  The  five  place 
table  is  perhaps  the  most  popular  of  all. 

Referring  now  to  the  four  place  table,  to 
find  a common  logarithm  of  a number,  pro- 
ceed as  follows.  Suppose  the  number  is  5576, 
First,  determine  the  characteristic.  An  inspec- 
tion will  show  that  the  characteristic  should 
be  3.  This  figure  is  placed  to  the  left  of  the 
decimal  point.  The  mantissa  is  now  found  by 
reference  to  the  logarithm  table.  The  first  two 
numbers  are  55;  glance  down  the  N column 
until  coming  to  these  figures.  Advance  to  the 
tight  until  coming  in  line  with  the  column 
headed  7;  the  mantissa  will  be  7459.  (Note  that 
the  column  headed  7 corresponds  to  the  third 
figure  in  the  number  5576.)  Place  the  mantissa 
7459  to  the  right  of  the  decimal  point,  making 
the  logarithm  of  5576  now  read  3.7459.  Impor- 
tant: do  not  consider  the  last  figure  6 in  the 


760  Radio  Mathematics  and  Calculations  THE  RADIO 


N 

L. 

0* 

1 

2 

3 

4 

5 

6 

7 

8 

9 

P.P. 

2S0 

39 

794 

81  i 

829 

846 

863 

881 

898 

?I5 

933 

950 

SI 

m 

985 

•002' 

•019 

•037 

•054 

•071 

•098 

•106’ 

•123 

18 

252 

40 

140 

157 

175 

192 

209 

226 

243 

2«l 

278 

295 

1 1.5 

253 

312 

329 

346 

364 

381 

398 

415 

432 

449 

466 

2 3.i 

254 

483 

500 

518 

535 

552 

569 

586 

M3 

620 

637 

3 5.4 

4 72 

255 

654 

671 

688 

705 

722 

739 

756 

773 

790 

807 

•tc. 

Figure  6. 

A SMALL  SECTION  OF  A FIVE  PLACE 
LOGARITHM  TABLE. 

Logarithms  may  be  fourtd  with  greater  aeeu~ 
racy  with  such  tables,  but  they  are  only  of 
use  when  the  accuracy  of  the  original  data 
warrants  greater  precision  in  the  figure  work. 
Slightly  greater  accuracy  may  be  obtained  for 
Intermediate  points  by  interpolation,  as  ex- 
gained  In  the  text. 


number  5576  when  looking  for  the  mantissa 
in  the  accompanying  four  place  tables;  in 
fact,  one  may  usually  disregard  all  digits  be- 
yond the  first  three  when  determining  the  man- 
tissa. (Interpolation,  sometimes  used  to  find  a 
logarithm  more  accurately,  is  unnecessary  un- 
less warranted  by  unusual  accuracy  in  the 
available  data.)  However,  be  doubly  sure  to 
include  all  figures  when  ascertaining  the  mag- 
nitude of  the  characteristic. 

To  find  the  anti-logarithm,  the  table  is  used 
in  reverse.  As  an  example,  let  us  find  the  anti- 
logarithm of  1.272  or,  in  other  words,  find 
the  number  of  which  1.272  is  the  logarithm. 
Look  in  the  table  for  the  mantissa  closest  to 
272.  This  is  found  in  the  first  half  of  the  table 
and  the  nearest  value  is  2718.  Write  down  the 
first  two  significant  figures  of  the  anti-loga- 
rithm by  taking  the  figures  at  the  beginning  of 
the  line  on  which  2718  was  found.  This  is  18; 
add  to  this,  the  digit  above  the  column  in 
which  2718  was  found;  this  is  7.  The  anti-log- 
arithm is  187  but  we  have  not  yet  placed  the 
decimal  point.  The  characteristic  is  1,  which 
means  that  there  should  be  two  digits  to  the 
left  of  the  decimal  point.  Hence,  18.7  is  the 
anti-logarithm  of  1.272. 

For  the  sake  of  completeness  we  shall  also 
describe  the  same  operation  with  a five-place 
table  where  interpolation  is  done  by  means  of 
tables  of  proportional  parts  (P.P.  tables). 
Therefore  we  are  reproducing  here  a small 
part  of  one  page  of  a five-place  table. 

Finding  the  logarithm  of  0.025013  is  done  as 
follow's:  We  can  begin  with  the  characteristic, 
which  is  —2.  Next  find  the  first  three  digits  in 
the  column,  headed  by  N and  immediately 
after  this  we  see  39,  the  first  two  digits  of  the 
mantissa.  Then  look  among  the  headings  of 
the  other  columns  for  the  next  digit  of  the 
number,  in  this  case  1.  In  the  column,  headed 
by  1 and  on  the  line  headed  250,  we  find  the 
next  three  digits  of  the  logarithm,  811.  So  far. 


the  logarithm  is  —2.39811  but  this  is  the  loga- 
rithm of  0.025010  and  we  want  the  logarithm 
of  0.025013.  Here  we  can  interpolate  by  ob- 
serving that  the  difference  between  the  log  of 

O. 02501  and  0.02502  is  829  - 811  or  18,  in 
the  last  two  significant  figures.  Looking  in  the 

P. P.  table  marked  18  we  find  after  3 the  num- 
ber 5.4  which  is  to  be  added  to  the  logarithm. 

-2.39811 

5.4 


— 2.39816,  Hie  logorithm  of  0.025013 

Since  our  table  is  only  good  to  five  places, 
we  must  eliminate  the  last  figure  given  in  the 
P.P.  table  if  it  is  less  than  5,  otherwise  we 
must  add  one  to  the  next  to  the  last  figure, 
rounding  off  to  a whole  number  in  the  P.P. 
table. 

Finding  the  anti-logarithm  is  done  the  same 
way  but  with  the  procedure  reversed.  Suppose 
it  is  required  to  find  the  anti-logarithm  of 
0.40100.  Find  the  first  two  digits  in  the  column 
headed  by  L.  Then  one  must  look  for  the  next 
three  digits  or  the  ones  nearest  to  it,  in  the 
columns  after  40  and  on  the  lines  from  40  to 
41.  Now  here  we  find  that  numbers  in  the 
neighborhood  of  100  occur  only  with  an  aster- 
isk on  the  line  just  before  40  and  still  after  39. 
The  asterisk  means  that  instead  of  the  39  as 
the  first  two  digits,  these  mantissas  should 
have  40  as  the  first  two  digits.  The  logarithm 
0.40100  is  between  the  logs  0.40088  and 
0.40106;  the  anti-logarithm  is  between  2517 
and  2518.  The  difference  between  the  two 
logarithms  in  the  table  is  again  18  in  the 
last  two  figures  and  our  logarithm  0.40100 
differs  with  the  lower  one  12  in  the  last 
figures.  Look  in  the  P.P.  table  of  18  which 
number  comes  closest  to  12.  This  is  found 
to  be  12.6  for  7 x 1.8  = 12.6.  Therefore 
we  may  add  the  digit  7 to  the  anti-logarithm 
already  found;  so  we  have  25177.  Next, 
place  the  decimal  point  according  to  the  rules: 
There  are  as  many  digits  to  the  left  of  the 
decimal  point  as  indicated  in  the  characteris- 
tic plus  one.  The  anti-logarithm  of  0.40100  is 
2.5177. 

In  the  following  examples  of  the  use  of  log- 
arithms we  shall  use  only  three  places  from  the 
tables  printed  in  this  chapter  since  a greater 
degree  of  precision  in  our  calculations  would 
not  be  warranted  by  the  accuracy  of  the  data 
given. 

In  a 375  ohm  bias  resistor  flows  a current 
of  41.5  milliamperes;  how  many  watts  are  dis- 
sipated by  the  resistor.^ 

We  write  the  equation  for  power  in  watts: 

P=  PR 


HANDBOOK 


The  Decibel  761 


and  filling  in  the  quantities  in  question,  we 
have: 

P = 0.0415“  X 375 

Taking  logarithms, 

log  P = 2 log  0.0415  + log  375 

log  0.0415  = -2.618 

So  2 X log  0.0415  = -3.236 

log  375  = 2.574 

log  P = -1.810 

anfilog  = 0.646.  Answer  = 0.646  woHs 

Caution:  Do  not  forget  that  the  negative 
sign  before  the  characteristic  belongs  to  the 
characteristic  only  and  that  mantissas  are  al- 
ways positive.  Therefore  we  recommend  the 
other  notation,  for  it  is  less  likely  to  lead  to 
errors.  The  work  is  then  written: 

log  0.0415=  8.618-10 
2 X log  0.0415  = 17.236-20  = 7.236-10 
log  375  = 2.574 

logP  =9.810-10 

Another  example  follows  which  demon- 
strates the  ease  in  handling  powers  and  roots. 
Assume  an  all-wave  receiver  is  to  be  built, 
covering  from  550  kc.  to  60  me.  Can  this  be 
done  in  five  ranges  and  what  will  be  the  re- 
quired tuning  ratio  for  each  range  if  no  over- 
lapping is  required?  Call  the  tuning  ratio  of 
one  band,  x.  Then  the  total  tuning  ratio  for 
five  such  bands  is  x‘.  But  the  total  tuning  ratio 
for  all  bands  is  60/0.55.  Therefore: 


Taking  logarithms: 

, log  60  — log  0.55 

log  X = p-* 

log  60  1 .778 

log  0.55  -1.740 

subtract 

2.038 

Remember  again  that  the  mantissas  are  posi- 
tive and  the  characteristic  alone  can  be  nega- 
tive. Subtracting  —1  is  the  same  as  adding  +1. 

log  X = = 0.408 

X = onlilog  0.408  = 2.5^ 

The  tuning  ratio  should  be  2.56. 


Power 

db 

Ratio 

0 

1.00 

1 

1.26 

2 

1.58 

3 

2.00 

4 

2.51 

$ 

3.16 

6 

3.98 

7 

5.01 

8 

6.31 

9 

7.94 

10 

10.00 

20 

100 

30 

1,000 

40 

10,000 

50 

100,000 

60 

1 ,000,000 

70 

1 0,000,000 

80 

100,000,000 

A TABLE 

Figure  7. 

OF  DECIBEL  GAINS  VERSUS 
POWER  RATIOS. 

The  Decibel 


The  decibel  is  a unit  for  the  comparison  of 
power  or  voltage  levels  in  sound  and  electrical 
work.  The  sensation  of  our  ears  due  to  sound 
waves  in  the  surrounding  air  is  roughly  pro- 
portional to  the  logarithm  of  the  energy  of  the 
sound-wave  and  not  proportional  to  the  energy 
itself.  For  this  reason  a logarithmic  unit  is  used 
so  as  to  approach  the  reaction  of  the  ear. 

The  decibel  represents  a ratio  of  two  power 
levels,  usually  connected  with  gains  or  loss  due 
to  an  amplifier  or  other  network.  The  decibel 
is  defined 

N..  = lOlog-^ 

where  Po  stands  for  the  output  power,  Pi  for 
the  input  power  and  Nat  for  the  number  of 
decibels.  When  the  answer  is  positive,  there  is 
a gain;  when  the  answer  is  negative,  there  is 
a loss. 

The  gain  of  amplifiers  is  usually  given  in 
decibels.  For  this  purpose  both  the  input  power 
and  output  power  should  be  measured.  Ex- 
ample: Suppose  that  an  intermediate  amplifier 
is  being  driven  by  an  input  power  of  0.2  watt 
and  after  amplification,  the  output  is  found  to 
be  6 watts. 


log  30  = 1.48 


Therefore  the  gain  is  10  X 1.48  = 14.8 
decibels.  The  decibel  is  a logarithmic  unit; 
when  the  power  was  multiplied  by  30,  the 
power  level  in  decibels  was  increased- by  14.8 
decibels,  or  14.8  decibels  added. 


762  Radio  Mathematics  and  Calculations  THE  RADIO 


TUBE  STEP-UP 

CAIN  = A RATIO=  3.5:1 


The  voltage  gain  in  decibels  in  this  stage  is 
equal  to  the  amplification  in  the  tube  plus  the 
step-up  ratio  of  the  transformer,  both  ex- 
pressed in  decibels. 


When  one  amplifier  is  to  be  followed  by 
another  amplifier,  power  gains  are  multiplied 
but  the  decibel  gains  are  added.  If  a main  am- 
plifier having  a gain  of  1,000,000  (power  ratio 
is  1,000,000)  is  preceded  by  a pre-amplifier 
with  a gain  of  1000,  the  total  gain  is  1,000,- 
000,000.  But  in  decibels,  the  first  amplifier  has 
a gain  of  60  decibels,  the  second  a gain  of  30 
decibels  and  the  two  of  them  will  have  a gain 
of  90  decibels  when  connected  in  cascade. 
(This  is  true  only  if  the  two  amplifiers  are 
properly  matched  at  the  junction  as  otherwise 
there  will  be  a reflection  loss  at  this  point 
which  must  be  subtracted  from  the  total.) 

Conversion  of  power  ratios  to  decibels  or 
vice  versa  is  easy  with  the  small  table  shown 
on  these  pages.  In  any  case,  an  ordinary  loga- 
rithm table  will  do.  Find  the  logarithm  of  the 
power  ratio  and  multiply  by  ten  to  find  deci- 
bels. 

Sometimes  it  is  mote  convenient  to  figure 
decibels  from  voltage  or  current  ratios  or  gains 
rather  than  from  power  ratios.  This  applies 
especially  to  voltage  amplifiers.  The  equation 
for  this  is 

Ndb  = 20  log  or  20  log 

where  the  subscript,  „,  denotes  the  output  volt- 
age or  current  and  i the  input  voltage  or  cur- 
rent. Remember,  this  equation  is  true  only  if 
the  voltage  or  current  gain  in  question  repre- 
sents a power  gain  which  is  the  square  of  it 
and  not  if  the  power  gain  which  results  from 
this  is  some  other  quantity  due  to  impedance 
changes.  This  should  be  quite  clear  when  we 
consider  that  a matching  transformer  to  con- 
nect a speaker  to  a line  or  output  tube  does 
not  represent  a gain  or  loss;  there  is  a voltage 
change  and  a current  change  yet  the  power  re- 
mains the  same  for  the  impedance  has  changed. 

On  the  other  hand,  when  dealing  with  volt- 
age amplifiers,  we  can  figure  the  gain  in  a 
stage  by  finding  the  voltage  ratio  from  the  grid 
of  the  first  tube  to  the  grid  .of  the  next  tube. 


Example:  In  the  circuit  of  Figure  8,  the  gain 
in  the  stage  is  equal  to  the  amplification  in  the 
tube  and  the  step-up  ratio  of  the  transformer. 
If  the  amplification  in  the  tube  is  10  and  the 
step-up  in  the  transformer  is  3.5,  the  voltage 
gain  is  35  and  the  gain  in  decibels  is: 

20  X log  35  = 20  X 1.54  = 30.8  db 

Decibels  os  The  original  use  of  the  decibel 

Power  Level  was  only  as  a ratio  of  power 

levels — not  as  an  absolute 
measure  of  power.  However,  one  may  use  the 
decibel  as  such  an  absolute  unit  by  fixing  an 
arbitrary  "zero”  level,  and  to  indicate  any 
power  level  by  its  number  of  decibels  above  or 
below  this  arbitrary  zero  level.  This  is  all  very 
good  so  long  as  we  agree  on  the  zero  level. 

Any  power  level  may  then  be  converted  to 
decibels  by  the  equation: 

Ndb  = 10  log  5^ 

“ref. 

where  N.ib  is  the  desired  power  level  in  deci- 
bels, Po  the  output  of  the  amplifier,  Pret.  the 
arbitrary  reference  level. 

The  zero  level  most  frequently  used  (but  not 
always)  is  6 milliwatts  or  0.006  watts.  For  this 
zero  level,  the  equation  reduces  to 

N„b=  lOlogj^, 

Example:  An  amplifier  using  a 6F6  tube 
should  be  able  to  deliver  an  undistorted  output 
of  3 watts.  How  much  is  this  in  decibels? 


10  X log  500  = 10  X 2.70  = 27.0 

Therefore  the  power  level  at  the  output  of 
the  6F6  is  27.0  decibels.  When  the  power  level 
to  be  converted  is  less  than  6 milliwatts,  the 
level  is  noted  as  negative.  Here  we  must  re- 
member all  that  has  been  said  regarding  loga- 
rithms of  numbers  less  than  unity  and  the  fact 
that  the  characteristic  is  negative  but  not  the 
mantissa. 

A preamplifier  for  a microphone  is  feeding 
1.5  milliwatts  into  the  line  going  to  the  regu- 
lar speech  amplifier.  What  is  this  power  level 
expressed  in  decibels? 

decibels  = 10  logo;^= 

’0  = ’O'og  0-25 

Log  0.25  = —1.398  (from  table).  There- 
fore, 10  X -1.398  = (10  X -1  = -10) 
+ (10  X .398  = 3.98);  adding  the  products 
algebraically,  gives  —6.02  db. 

The  conversion  chart  reproduced  in  this 
chapter  will  be  of  use  in  converting  decibels  to 
watts  and  vice  versa. 


HANDBOOK 


Decibel-Power  Conversion  763 


Figui«  9. 

CONVERSION  CHART:  POWER  TO  DECIBELS 

Power  /eve/s  between  6 m/crom/crowotts  an<l  6000  wotts  moy  be  reterred  to  corresponding  decibel 
levels  between  >90  ond  60  db,  and  vice  versa,  by  means  of  the  above  chart.  Fifteen  ranges  are 
provided,  Bach  curve  begins  at  the  same  point  where  the  preceding  one  ends,  enabling  uninterrupted 
coverage  of  the  wide  db  and  power  ranges  with  condensed  chart.  For  examine:  the  lowermoet  curve 
ends  at  >00  db  or  60  micromicrowatts  and  the  next  range  starts  at  the  same  leveL  Zero  db  level  is 
taken  at  6 milliwaitt  (,006  wait). 


LEVEL  IN  D.B. 


764  Radio  Mathematics  and  Calculations  THE  RADIO 


Converting  Decibels  It  is  often  convenient 

to  Power  to  be  able  to  convert  a 

decibel  value  to  a pow- 
er equivalent.  The  formula  used  for  this  oper- 
ation is 

P = 0.006  X ontilog  ^ 

where  P is  the  desired  level  in  watts  and  Ndb 
the  decibels  to  be  converted. 

To  determine  the  power  level  P from  a dec- 
ibel equivalent,  simply  divide  the  decibel  value 
by  10;  then  take  the  number  comprising  the 
antilog  and  multiply  it  by  0.006;  the  product 
gives  the  level  in  watts. 

Nofe:  In  problems  dealing  with  the  conver- 
sion of  minus  decibels  to  power,  it  often  hap- 
pens that  the  decibel  value  — Nds  is  not 
divisible  by  10.  When  this  is  the  case, 

the  numerator  in  the  factor  — must  be 

made  evenly  divisible  by  10,  the  negative  signs 
must  be  observed,  and  the  quotient  labeled  ac- 
cordingly. 

To  make  the  numerator  evenly  divisible  by 
10  proceed  as  follows:  Assume,  for  example, 
that  —Ndb  is  some  such  value  as  —38;  to 
make  this  figure  evenly  divisible  by  10,  we 
must  add  —2  to  it,  and,  since  we  have  added 
a negative  2 to  it,  we  must  also  add  a positive 
2 so  as  to  keep  the  net  result  the  same. 

Our  decibel  value  now  stands,  —40  + 2. 
Dividing  both  of  these  figures  by  10,  as  in  the 
equation  above,  we  have  — 4 and  4-0.2.  Put- 
ting the  two  together  we  have  the  logarithm 
— 4.2  with  the  negative  characteristic  and  the 
positive  mantissa  as  required. 

The  following  examples  will  show  the  tech- 
nique to  be  followed  in  practical  problems. 

(a)  The  output  of  a certain  device  is  rated 
at  —74  db.  What  is  the  power  equivalent? 
Solution: 

( not  evenly  divisible  by  1 0 ) 

Routine: 

-74 

- 6 -P6 

-80  4-6 

Ndb  — 80  -f-  6 Q g, 

Jo= — 10 — - 
ontilog  —8.8  = 0.000  000  04 
.006  X 0.000  000  04  = 

0.000  000  000  24  watt  or 
240  micro-microwatt 

(b)  This  example  differs  somewhat  from 
that  of  the  foregoing  one  in  that  the  mantissas 
are  added  differently.  A low-powered  amplifier 
has  an  input  signal  level  of  —17.3  db.  How 
many  milliwatts  does  this  value  represent? 


Solution; 

-17.3 

- 2.7  + 2.7 

-20  + 2.7 

Ndb  __  -20  + 2.7 

Antilog  -2.27  = 0.0186 

0.006  X 0.0186  = 0.000  1116  watt  or 
0.1 1 1 6 milliwatt 

Input  voltages:  To  determine  the  required 
input  voltage,  take  the  peak  voltage  necessary 
to  drive  the  last  class  A amplifier  tube  to  max- 
imum output,  and  divide  this  figure  by  the  to- 
tal overall  voltage  gain  of  the  preceding  stages. 

Computing  Specifications:  From  the  preced- 
ing explanations  the  following  data  can  be 
computed  with  any  degree  of  accuracy  war- 
ranted by  the  circumstances; 

(1)  Voltage  amplification 

(2)  Overall  gain  in  db 

(3)  Output  signal  level  in  db 

(4)  Input  signal  level  in  db 

(5)  Input  signal  level  in  watts 

(6)  Input  signal  voltage 

When  a power  level  is  available  which  must 
be  brought  up  to  a new  power  level,  the  gain 
required  in  the  intervening  amplifier  is  equal 
to  the  difference  between  the  two  levels  in  dec- 
ibels. If  the  required  input  of  an  amplifier  for 
full  output  is  —30  decibels  and  the  output 
from  a device  to  be  used  is  but  —45  decibels, 
the  pre-amplifier  required  should  have  a gain 
of  the  difference,  or  15  decibels.  Again  this  is 
true  only  if  the  two  amplifiers  are  properly 
matched  and  no  losses  are  introduced  due  to 
mismatching. 

Push-Pull  To  double  the  output  of  any  cas- 
Amplifiers  cade  amplifier,  it  is  only  neces- 
sary to  connect  in  push-pull  the 
last  amplifying  stage,  and  replace  the  inter- 
stage and  output  transformers  with  push-pull 
types. 

To  determine  the  voltage  gain  (voltage  ra- 
tio) of  a push-pull  amplifier,  take  the  ratio  of 
one  half  of  the  secondary  winding  of  the  push- 
pull  transformer  and  multiply  it  by  the  n of 
one  of  the  output  tubes  in  the  push-pull  stage; 
the  product,  when  doubled,  will  be  the  voltage 
amplification,  or  step-up. 

Other  Units  and  When  working  with  deci- 
Zero  Levels  bels  one  should  not  im- 

mediately take  for  granted 
that  the  zero  level  is  6 milliwatts  for  there  are 
other  zero  levels  in  use. 

In  broadcast  stations  an  entirely  new  system 
is  now  employed.  Measurements  made  in 


HANDBOOK 


Trigonometry  765 


acoustics  are  now  made  with  the  standard  zero 
level  of  lO"'®  watts  per  square  cm. 

Microphones  are  often  rated  with  reference 
to  the  following  zero  level:  one  volt  at  open 
circuit  when  the  sound  pressure  is  one  millibar. 
In  any  case,  the  rating  of  the  microphone  must 
include  the  loudness  of  the  sound.  It  is  obvious 
that  this  zero  level  does  not  lend  itself  readily 
for  the  calculation  of  required  gain  in  an  am- 
plifier. 

The  VU:  So  far,  the  decibel  has  always  re- 
ferred to  a type  of  signal  which  can  readily  be 
measured,  that  is,  a steady  signal  of  a single 
frequency.  But  what  would  be  the  power  level 
of  a signal  which  is  constantly  varying  in  vol- 
ume and  frequency?  The  measurement  of  volt- 
age would  depend  on  the  type  of  instrument 
employed,  whether  it  is  measured  with  a 
thermal  square  law  meter  or  one  that  shows 
average  values;  also,  the  inertia  of  the  move- 
ment will  change  its  indications  at  the  peaks 
and  valleys. 

After  considerable  consultation,  the  broad- 
cast chains  and  the  Bell  System  have  agreed 
on  the  VU.  The  level  in  VU  is  the  level  in 
decibels  above  1 milliwatt  zero  level  and  meas- 
ured with  a carefully  defined  type  of  instru- 
ment across  a 600  ohm  line.  So  long  as  we 
deal  with  an  unvarying  sound,  the  level  in  VU 
is  equal  to  decibels  above  1 milliwatt;  but 
when  the  sound  level  varies,  the  unit  is  the 
VU  and  the  special  meter  must  be  used.  There 
is  then  no  equivalent  in  decibels. 

The  Neper;  We  might  have  used  the  natural 
logarithm  instead  of  the  common  logarithm 
when  defining  our  logarithmic  unit  of  sound. 
This  was  done  in  Europe  and  the  unit  obtained 
is  known  as  the  neper  or  napier.  It  is  still 
found  in  some  American  literature  on  filters. 

1 neper  = 8.686  decibels 

1 decibel  = 0.1151  neper 

AC  Meters  With  Many  test  instruments 
Decibel  Scales  ate  now  equipped  with 

scales  calibrated  in  deci- 
bels which  is  very  handy  when  making  meas- 
urements of  frequency  characteristics  and  gain. 
These  meters  are  generally  calibrated  for  con- 
nection across  a 500  ohm  line  and  for  a zero 
level  of  6 milliwatts.  When  they  are  connected 
across  another  impedance,  the  reading  on  the 
meter  is  no  longer  correct  for  the  zero  level  of 
6 milliwatts.  A correction  factor  should  be 
applied  consisting  in  the  addition  or  subtrac- 
tion of  a steady  figure  to  all  readings  on  the 
meter.  This  figure  is  given  by  the  equation; 

db  to  be  added  =10  log  ^ 

where  Z is  the  impedance  of  the  circuit  under 
measurement. 


Figure  10. 

THE  CIRCLE  IS  DIVIDED  INTO 
FOUR  QUADRANTS  BY  TWO  PER- 
PENDICULAR LINES  AT  RIGHT 
ANGLES  TO  EACH  OTHER. 

Th0  northeast**  quadrant  thus  formed  is 
known  as  the  first  quadrant;  the  others  are 
numbered  consecutively  in  a counterclockwise 
direction. 


Trigonometry 

Definition  Trigonometry  is  the  science  of 
and  Use  mensuration  of  triangles.  At  first 
glance  triangles  may  seem  to 
have  little  to  do  with  electrical  phenomena; 
however,  in  a.c.  work  most  currents  and  volt- 
ages follow  laws  equivalent  to  those  of  the 
various  trigonometric  relations  which  we  are 
about  to  examine  briefly.  Examples  of  their 
application  to  a.c.  work  will  be  given  in  the 
section  on  Vectors. 

Angles  ate  measured  in  degrees  or  in  radi- 
ans. The  circle  has  been  divided  into  360 
degrees,  each  degree  into  60  minutes,  and  each 
minute  into  60  seconds.  A decimal  division  of 
the  degree  is  also  in  use  because  it  makes  cal- 
culation easier.  Degrees,  minutes  and  seconds 
are  indicated  by  the  following  signs;  ' and  ". 
Example:  6°  5'  23"  means  six  degrees,  five 
minutes,  twenty-three  seconds.  In  the  decimal 
notation  we  simply  write  8.47°,  eight  and 
forty-seven  hundredths  of  a degree. 

When  a circle  is  divided  into  four  quadrants 
by  two  perpendicular  lines  passing  through 
the  center  (Figure  10)  the  angle  made  by  the 
two  lines  is  90  degrees,  known  as  a right  angle. 
Two  tight  angles,  or  180°  equals  a straight 
angle. 

The  radian;  If  we  take  the  radius  of  a circle 
and  bend  it  so  it  can  covet  a part  of  the  cir- 
cumference, the  arc  it  covets  subtends  an  angle 
called  a radian  (Figure  11).  Since  the  circum- 
ference of  a circle  equals  2 times  the  radius, 
there  are  Iv  radians  in  360°.  So  we  have  the 
following  relations: 

lradian  = 57°  17' 45"=  57.2958°  ■77  = 3.14159 

1 degree =0.01 745  radians 
17  radians  = 180°'  ■77/2  radians =90° 

ir/3  radians  = 60° 


766  Radio  Mathematics  and  Calculations  THE  RADIO 


THE  RADIAN. 

A radian  is  an  angle  whose  arc  is  exactly  equal 
to  the  length  of  either  side.  Note  that  the 
angle  is  constant  regardless  of  the  length  of 
the  side  and  the  arc  so  loirg  as  they  are  equal. 
A radian  equals  57.2958°. 


In  trigonometry  we  consider  an  angle  gen- 
erated by  two  lines,  one  stationary  and  the 
other  rotating  as  if  it  were  hinged  at  0,  Figure 
12.  Angles  can  be  greater  than  180  degrees  and 
even  greater  than  360  degrees  as  illustrated  in 
this  figure. 

Two  angles  are  complements  of  each  other 
when  their  sum ’is  90°.  or  a right  angle.  A is 
the  complement  of  B and  B is  the  complement 
of  A when 

A=(90°-B) 
and  when 
B=(90‘’-A) 

Two  angles  are  supplements  of  each  other 
when  their  sum  is  equal  to  a straight  angle,  or 
180°.  A is  the  supplement  of  B and  B is  the 
supplement  of  A when 


In  the  angle  A,  Figure  13 A,  a line  is  drawn 
from  P,  perpendicular  to  b.  Regardless  of  the 
point  selected  for  P,  the  ratio  a/c  will  always 
be  the  same  for  any  given  angle,  A.  So  will  all 
the  other  proportions  between  a,  h,  and  c re- 
main constant  regardless  of  the  position  of 
point  P on  c.  The  six  possible  ratios  each  are 
named  and  defined  as  follows: 


sine  A = 


a 

c 


cosine  A — — 
c 


tangent  A = r- 


cotangent  A 


b 

a 


secant  A = 


c 

b 


cosecant  A = — 
a 


Let  us  take  a special  angle  as  an  example. 
For  instance,  let  the  angle  A be  60  degrees  as 
in  Figure  13B.  Then  the  relations  between  the 
sides  are  as  in  the  figure  and  the  six  functions 
become: 


sin.  60° 


a 

c 


V2VI 

1 


1/2 


cos  60° 


tan  60°  = ^ - - /3 


cot  60°  = 


1 


1/2  _L 

VI 


sec  60°  b ~ 1/2  ^ 


CSC  60®  = “ 
a 


V2V1 


= Vi  VI 


=2/3/3 


A=(180‘’-B) 

and 

B=(180°-A) 


Another  example:  Let  the  angle  be  45°,  then 
the  relations  between  the  lengths  of  a,  b,  and 
c ate  as  shown  in  Figure  13C,  and  the  six 
functions  are: 


A 


AN  ANGLE  IS  GENERATED  BY  TWO  LINES,  ONE  STATIONARY 

ROTATING. 


AND  THE  OTHER 


The  line  OX  is  stationary;  the  line  with  the  small  arrow  at  the  far  end  rotates  in  a counterclockwise 
direction.  At  the  position  illustrated  in  the  lefthandmosi  section  of  the  drawing  it  makes  an  angle, 
A,  which  is  less  than  90°  and  is  therefore  in  the  first  quadrant.  In  the  position  shown  in  the  second 
portion  of  the  drawing  the  angle  A has  increased  to  such  a value  that  it  now  lies  in  the  third 
quadrant;  note  that  an  angle  can  be  greater  than  180°.  In  the  third  illustration  the  angle  A is  in 
the  fourth  quadrant.  In  the  fourth  position  the  rotating  vector  has  made  more  than  one  complete 
revolution  and  is  hence  in  the  fifth  quadrant;  since  the  fifth  quadrant  is  an  exact  repetition  of  the 
first  quadrant,  its  values  will  be  the  same  as  in  the  lefthandmost  portion  of  the  illustration. 


768  Radio  Mathematics  and  Calculations  THE  RADIO 


B 


Figure  15. 

In  this  figure  ihe  sides  a,  b,  and  c are  used 
to  define  the  trigonometric  functions  of  angle 
B as  well  as  angle  A. 


sin  B = b/c 
cos  B = a/c 
tan  B = b/a 
cot  B = a/b 


b = c sin  B 
a = c cos  B 
b = a tan  B 
a = b cot  B 


Functions  of  Angles  In  angles  greater  than 

Greater  than  90  degrees,  the  values 

90  Degrees  of  a and  i become  neg- 

ative on  occasion  in  ac- 
cordance with  the  rules  of  Cartesian  coordi- 
nates. When  i is  measured  from  0 towards 
the  left  it  is  considered  negative  and  similarly, 
when  a is  measured  from  0 downwards,  it  is 
negative.  Referring  to  Figure  16,  an  angle  in 
the  second  quadrant  (between  90°  and  180°) 
has  some  of  its  functions  negative; 

sm  A = = pos. 

. a 

tan  A = = neg. 

sec  A = = neg. 

For  an  angle  in  the  third  quadrant  (180°  to 
270°),  the  functions  are 

— a ~b 

sin  A = -g-  = neg.  cos  A — neg. 

— a A 

tan  A = = pos.  cot  A = = pos. 


cos  A = — = neg. 
A -b 

cot  A = — = neg. 
cosec  A = -g—  = pos. 


POSITIVE  FUNCTIONS 


SECOND  OUADRANT 


FIRST  OUADRANT 


eecec 


all  function! 


tan,  cot 


cosine,  secant 


THIRD  QUADRANT 


FOURTH  QUADRANT 


Figure  17. 

SIGNS  OF  THE  TRIGONOMETRIC 
FUNCTIONS. 

The  functions  listed  in  this  diagram  are  posi~ 
five;  all  other  functions  are  negative. 


C C 

sec  A = ^ = neg.  cosec  A = •—  = neg. 

D 2 

And  in  the  fourth  quadrant  (270°  to  360°) : 

. . -a_  b 

sm  A = -g-  = neg.  cos  A = -g-  = pos. 

A — ® A b 

tan  A = -g-  = neg.  cot  A = = neg. 

sec  A = -g-  = pos.  cosec  A = neg. 

Summarizing,  the  sign  of  the  functions  in 
each  quadrant  can  be  seen  at  a glance  from 
Figure  17,  where  in  each  quadrant  are  written 
the  names  of  functions  which  are  positive; 
those  not  mentioned  are  negative. 


TRIGONOMETRIC  FUNCTIONS  IN  THE  SECOND,  THIRD,  AND  FOURTH  QUADRANTS. 

TAe  trigonometric  functions  in  these  quadrants  are  similar  to  first  quadrant  valuos,  but  tAe 
signs  of  tho  functions  yary  as  listod  in  the  text  and  in  Figuro  17 e 


HANDBOOK 


Trigonometric  Curves  769 


(A) 


Figure  18. 

SINE  AND  COSINE  CURVES. 

In  (A)  we  have  a sine  curve 
drawn  in  Cartesian  coordinates. 
This  is  the  usual  representation 
of  an  alternating  current  wave 
without  substantial  harmonics.  In 
(B)  we  have  a cosine  wave; 
note  that  it  is  exactly  similar 
to  a sine  wave  displaced  by 
90°  or  tt/2  radians. 


(B) 


Graphs  of  Trigono-  The  sine  wave.  When 
metric  Functions  we  have  the  relation 

•)  — sinx,  where  x is  an 
angle  measured  in  radians  or  degrees,  we  can 
draw  a curve  of  y versus  .r  for  all  values  of 
the  independent  variable,  and  thus  get  a good 
conception  how  the  sine  varies  with  the  mag- 
nitude of  the  angle.  This  has  been  done  in 
Figure  18A.  We  can  learn  from  this  curve  the 
following  facts. 

1.  The  sine  varies  between  +1  and  —1 

2.  It  is  a periodic  curve,  repeating  itself  after 
every  multiple  of  2'7r  or  360° 

3.  Sin  X = sin  (180°  — x)  or  sin  (ir  — x) 

4.  Sin  X = —sin  (180°  + x),  or 
— sin  (tr  + x) 

The  cosine  wave.  Making  a curve  for  the 
function  y = cos  x,  we  obtain  a curve  similar 
to  that  for  y = sin  x except  that  it  is  displaced 
by  90°  or  rr/l  radians  with  respect  to  the 
Y-axis.  This  curve  (Figure  18B)  is  also  peri- 
odic but  it  does  not  start  with  zero.  We  read 
from  the  curve: 

1.  The  value  of  the  cosine  never  goes  be- 
yond + 1 or  —1 

2.  The  curve  repeats,  after  every  multiple 
of  2v7  radians  or  360° 


3.  Cos  X = —cos  (180°  — x)  or 

— cos  (w  — x) 

4.  Cos  X = cos  (360°— x)  or  cos  (2';r  — x) 

The  graph  of  the  tangent  is  illustrated  in 

Figure  19.  This  is  a discontinuous  curve  and 
illustrates  well  how  the  tangent  increases  from 
zero  to  infinity  when  the  angle  increases  from 
zero  to  90  degrees.  Then  when  the  angle  is 
further  increased,  the  tangent  starts  from 
minus  infinity  going  to  zero  in  the  second  quad- 
rant, and  to  infinity  again  in  the  third  quadrant. 

1.  The  tangent  can  have  any  value  between 
+ “ and  — “ 

2.  The  curve  repeats  and  the  period  is  it 
radians  or  180°,  not  2';r  radians 

3.  Tan  x = tan  (180°  +x)  or  tan  ('ir-fx) 

4.  Tan  x = —tan  (180°  — x)  or 

— tan  (tt  — x) 

The  graph  of  the  cotangent  is  the  inverse  of 
that  of  the  tangent,  see  Figure  20.  It  leads  us 
to  the  following  conclusions: 

1.  The  cotangent  can  have  any  value  be- 
tween + and  — ” 

2.  It  is  a periodic  curve,  the  period  being 
ST  radians  or  180° 

3.  Cot  x = cot  (180°  +x)  or  cot  (•tr  -fx) 

4.  Cot  x = —cot  (180°  — x)  or 

— cot  (tr  — x) 


Figure  19. 
TANGENT  CURVES. 


The  tangent  curve  increases  from  0 to  <x>  with 
an  angular  increase  of  90°.  In  the  next  780° 
ft  increases  from  — co  to  4*  ^ ■ 


Figure  20. 

COTANGENT  CURVES. 

Cotangent  curves  are  the  inverse  of  the  ton- 
gent  curves.  They  yery  from  -f  ot  to  — « in 
each  pair  of  quadrants. 


770  Radio  Mathematics  and  Calculations  THE  RADIO 


Figure  21 , 

ANOTHER  REPRESENTATION  OF 
TRIGONOMETRIC  FUNCTIONS. 

If  the  radius  of  a eirete  is  considered  as  the 
unit  of  measurement,  then  the  lengths  of  the 
various  lines  shown  in  this  diagram  are  numer- 
ically equal  to  the  functions  marked  adjacent 
to  them. 


The  graphs  of  the  secant  and  cosecant  are 
of  lesser  importance  and  will  not  be  shown 
here.  They  are  the  inverse,  respectively,  of  the 
cosine  and  the  sine,  and  therefore  they  vary 
from  +1  to  infinity  and  from  —1  to  —infinity. 

Perhaps  another  useful  way  of  visualiaing 
the  values  of  the  functions  is  by  considering 
Figure  21.  If  the  radius  of  the  circle  is  the  unit 
of  measurement  then  the  lengths  of  the  lines 
are  equal  to  the  functions  marked  on  them. 

Trigonometric  Tables  There  are  two  kinds  of 
trigonometric  tables. 
The  first  type  gives  the  functions  of  the  angles, 
the  second  the  logarithms ' of  the  functions. 
The  first  kind  is  also  known  as  the  table  of 
natural  trigonometric  functions. 

These  tables  give  the  functions  of  all  angles 
between  0 and  45°.  This  is  all  that  is  necessary 
for  the  function  of  an  angle  between  45°  and 
90°  can  always  be  written  as  the  co-function 
of  an  angle  below  45°.  Example;  If  we  had  to 
find  the  sine  of  48°,  we  might  write 
sin  48°=  cos  (90°  —48°)=  cos  42° 

Tables  of  the  logarithms  of  trigonometric 
functions  give  the  common  logarithms  (logw) 
of  these  functions.  Since  many  of  these  logar- 
ithms have  negative  characteristics,  one  should 
add  —10  to  all  logarithms  in  the  table  which 
have  a characteristic  of  6 or  higher.  For  in- 
stance, the  log  sin  24°  = 9-60931  —10.  Log 
tan  1°  = 8.24192  -10  but  log  cot  1°  = 
1.75808.  When  the  characteristic  shown  is  less 
than  6,  it  is  supposed  to  be  positive  and  one 
should  not  add  —10. 

Vectors 

A scalar  quantity  has  magnitude  only;  a 
vector  quantity  has  both  magnitude  and  direc- 
tion. When  we  speak  of  a speed  of  50  miles 
per  hour,  we  are  using  a scalar  quantity,  but 
when  we  say  the  wind  is  Northeast  and  has  a 


Vectors  moy  he  added  as  shown  in  these 
sketches.  In  each  case  the  long  vector  repre- 
sents the  vector  sum  of  the  smaller  vectors. 
For  many  engineering  applications  sufficient 
accuracy  can  be  obtained  by  this  method 
which  avoids  long  and  laborious  calculations. 


velocity  of  50  miles  per  hour,  we  speak  of  a 
vector  quantity. 

Vectors,  representing  forces,  speeds,  dis- 
placements, etc.,  are  represented  by  arrows. 
They  can  be  added  graphically  by  well  known 
methods  illustrated  in  Figure  22.  We  can  make 
the  parallelogram  of  forces  or  we  can  simply 
draw  a triangle.  The  addition  of  many  vectors 
can  be  accomplished  graphically  as  in  the  same 
figure. 

In  order  that  we  may  define  vectors  algebra- 
ically and  add,  subtract,  multiply,  or  divide 
them,  we  must  have  a logical  notation  system 
that  lends  itself  to  these  operations.  For  this 
purpose  vectors  can  be  defined  by  coordinate 
systems.  Both  the  Cartesian  and  the  polar  co- 
ordinates are  in  use. 

Vectors  Defined  Since  we  have  seen  how  the 
by  Cartesian  sum  of  two  vectors  is  ob- 
Coordinates  tained,  it  follows  from  Fig- 

ure 23,  that  the  vector  Z 
equals  the  sum  of  the  two  vectors  x and  y.  In 
fact,  any  vector  can  be  resolved  into  vectors 
along  the  X-  and  T-axis.  For  convenience  in 
working  with  these  quantities  we  need  to  dis- 


Figure  23. 

RESOLUTION  OF  VECTORS. 

Any  vector  such  os  Z may  be  resolved  into 
two  vectors,  x and  y,  along  the  X-  and  Y- 
axes.  If  vectors  are  to  be  added,  their  respec- 
tive X and  y components  may  be  added  to 
find  the  x and  y components  of  the  resultant 
vector. 


HANDBOOK 


Vectors  771 


ADDITION  OR  SUBTRACTION  OF 
VECTORS. 

Vectors  may  be  added  ar  subtracted  by 
adding  or  subtracting  their  x or  y com> 
ponents  separately. 


tinguish  between  the  X-  and  X-component, 
and  so  it  has  been  agreed  that  the  V-compo- 
nent  alone  shall  be  marked  with  the  letter  j. 
Example  (Figure  23): 

Z = 3 + 4i 

Note  again  that  the  sign  of  components 
along  the  X-axis  is  positive  when  measured 
from  0 to  the  right  and  negative  when  meas- 
ured from  0 towards  the  left.  Also,  the  compo- 
nent along  the  Y-axis  is  positive  when  meas- 
ured from  0 upwards,  and  negative  when 
measured  from  0 downwards.  So  the  vector, 
R,  is  described  as 

R = 5 - 3i 

Vector  quantities  are  usually  indicated  by 
some  special  typography,  especially  by  using  a 
point  over  the  letter  indicating  the  vector,  as 
R. 

Absolute  Volue  The  absolute  or  scalar 

of  0 Vector  value  of  vectors  such  as  Z 

or  R in  Figure  23  is  easily 
found  by  the  theorem  of  Pythagoras,  which 
states  that  in  any  right-angled  triangle  the 
square  of  the  side  opposite  the  tight  angle  is 
equal  to  the  sum  of  the  squares  of  the  sides 
adjoining  the  right  angle.  In  Figure  23,  OAB 
is  a right-angled  triangle;  therefore,  the  square 
of  OB  (or  Z)  is  equal  to  the  square  of  OA 
(or  x)  plus  the  square  of  AB  (or  y).  Thus  the 
absolute  values  of  Z and  R may  be  determined 
as  follows; 

|Z|  = ViMTV 

|Z|=  V 3^  + 4“  = 5 

I R 1 1=  V S'"  + 3=  = vT4  = 5.83 

The  vertical  lines  indicate  that  the  absolute 
or  scalar  value  is  meant  without  regard  to  sign 
or  direction. 


Addition  of  Vectors  An  examination  of  Fig- 
ure 24  will  show  that 

the  two  vectors 

R = xi  + j yi 

Z = xi  + iyi 

can  be  added,  if  we  add  the  X-components 
and  the  Y-components  separately. 

R + Z = xi  -h  X2  + i (yi  -f  y2) 

For  the  same  reason  we  can  carry  out  sub- 
traction by  subtracting  the  horizontal  compo- 
nents and  subtracting  the  vertical  components 

R — Z = X,  — X2  + 1 (yi  — y2) 

Let  us  consider  the  operator  If  we  have  a 
vector  a along  the  X-axis  and  add  a ; in  front 
of  it  (multiplying  by  ;)  the  result  is  that  the 
direction  of  the  vector  is  rotated  forward  90 
degrees.  If  we  do  this  twice  (multiplying  by 
/)  the  vector  is  rotated  forward  by  180  degrees 
and  now  has  the  value  —a.  Therefore  multi- 
plying by  f is  equivalent  to  multiplying  by  —1. 
Then 

j''  = — 1 ond  j = V — 1 

This  is  the  imaginary  number  discussed  be- 
fore under  algebra.  In  electrical  engineering 
the  letter  ; is  used  rather  than  i,  because  ; is 
already  known  as  the  symbol  for  current. 

Multiplying  Vectors  When  two  vectors  ate 
to  be  multiplied  we  can 
perform  the  operation  just  as  in  algebra,  re- 
membering that  j’  = —1. 

RZ  = (xi  -f  jyi)  (x2  -f-  jy2) 

= Xi  X2  + jxi  y2  + j X2  yi  + yi  y2 
= Xi  X2  — yi  y2  + i (xi  y2  + X2  yi) 

DivisioTi  has  to  be  carried  out  so  as  to  re- 
move the  /-term  from  the  denominator.  This 
can  be  done  by  multiplying  both  denominator 
and  numerator  by  a quantity  which  will  elimi- 
nate / from  the  denominator.  Example: 

^ Xi  -f  jyi l«i  + jyi)  (»2  — iy2) 

2 X2  + jyj  (X2  + jyz)  (xa  — jys) 

_xiX2  -4-  yiya  -b  j (x2yi  — Xiy2) 

“ X2’“  + y2“ 

Polar  Coordinates  A vector  can  also  be  de- 
fined in  polar  coordi- 
nates by  its  magnitude  and  its  vectorial  angle 
with  an  arbitrary  reference  axis.  In  Figure  25 


772  Radio  Mathematics  and  Calculations  THE  RADIO 


FiguK  25< 

IN  THIS  FIGURE  A VECTOR  HAS 
BEEN  REPRESENTED  IN  POLAR 
INSTEAD  OF  CARTESIAN  CO- 
ORDINATES. 

In  polar  coordinates  a vector  is  defined  by 
a magnitude  and  an  angle,  called  the  vec- 
torial angle,  instead  of  by  two  magnitudes 
as  in  Cartesian  coordinates. 


the  vector  Z has  a magnitude  50  and  a vector- 
ial angle  of  60  degrees.  This  will  then  be 
written 

Z = 50/60° 

A vector  a + jb  can  be  transformed  into 
polar  notation  very  simply  (see  Figure  26) 

Z = a -fjb  = \/P~+~P  /tan'“  | 

In  this  connection  tan'^  means  the  angle  of 
which  the  tangent  is.  Sometimes  the  notation 
arc  tan  b/a  \s  used.  Both  have  the  same  mean- 
ing. 

A polar  notation  of  a vector  can  be  trans- 
formed into  a Cartesian  coordinate  notation  in 
the  following  manner  (Figure  27) 

Z = p/A  = p cos  A -t-  jp  sin  A 

A sinusoidally  alternating  voltage  or  cur- 
rent is  symbolically  represented  by  a rotating 
vector,  having  a magnitude  equal  to  the  peak 
voltage  or  current  and  rotating  with  an  angular 
velocity  of  2'7rf  radians  per  second  or  as  many 
revolutions  per  second  as  there  are  cycles  per 
second. 

The  instantaneous  voltage,  e,  is  always  equal 
to  the  sine  of  the  vectorial  angle  of  this  rotat- 
ing vector,  multiplied  by  its  magnitude. 

e = E sin  ZitH 


Vectors  can  be  transformed  from  Cartesian 
into  polar  notation  as  shown  in  this  figure. 


which  flows  due  to  the  alternating  voltage  is 
not  necessarily  in  step  with  it.  The  rotating 
current  vector  may  be  ahead  or  behind  the 
voltage  vector,  having  a phase  difference  with 
it.  For  convenience  we  draw  these  vectors  as 
if  they  were  standing  still,  so  that  we  can  indi- 
cate the  difference  in  phase  or  the  phase  angle. 
In  Figure  28  the  current  lags  behind  the  volt- 
age by  the  angle  «,  or  we  might  say  that  the 
voltage  leads  the  current  by  the  angle  e. 

Vector  diagrams  show  the  phase  relations 
between  two  or  more  vectors  (voltages  and 
currents)  in  a circuit.  They  may  be  added  and 
subtracted  as  described;  one  may  add  a voltage 
vector  to  another  voltage  vector  or  a current 
vector  to  a current  vector  but  not  a current 
vector  to  a voltage  vector  ( for  the  same  reason 
that  one  cannot  add  a force  to  a speed).  Figure 
28  illustrates  the  relations  in  the  simple  series 
circuit  of  a coil  and  resistor.  We  know  that  the 
current  passing  through  coil  and  resistor  must 
be  the  same  and  in  the  same  phase,  so  we  draw 
this  current  I along  the  X-axis.  We  know  also 
that  the  voltage  drop  IR  across  the  resistor  is 
in  phase  with  the  current,  so  the  vector  IR  rep- 
resenting the  voltage  drop  is  also  along  the 
X-axis. 

The  voltage  across  the  coil  is  90  degrees 
ahead  of  the  current  through  it;  IX  must 
therefore  be  drawn  along  the  X-axis.  E the 
applied  voltage  must  be  equal  to  the  vectorial 
sum  of  the  two  voltage  drops,  IR  and  IX,  and 
we  have  so  constructed  it  in  the  drawing.  Now 
expressing  the  same  in  algebraic  notation,  we 
have 

E = IR  -f  jlX 
IZ  = IR  -f  jlX 

Dividing  by  I 

Z = R + iX 


The  alternating  voltage  therefore  varies  with 
time  as  the  sine  varies  with  the  angle.  If  we 
plot  time  horizontally  and  instantaneous  volt- 
age vertically  we  will  get  a curve  like  those 
in  Figure  18. 

In  alternating  current  circuits,  the  current 


Due  to  the  fact  that  a reactance  rotates  the 
voltage  vector  ahead  or  behind  the  current 
vector  by  90  degrees,  we  must  mark  it  with  a 
/ in  vector  notation.  Inductive  reactance  will 
have  a plus  sign  because  it  shifts  the  voltage 
vector  forwards;  a capacitive  reactance  is  neg- 


HANDBOOK 


ative  because  the  voltage  will  lag  behind  the 
current.  Therefore: 

X,  = + i 2s7fL 

„ 1 
~ ' 2iTfC 

In  Figure  28  the  angle  e is  known  as  the 
phase  angle  between  E and  I.  When  calculat- 
ing power,  only  the  real  components  count. 
The  power  in  the  circuit  is  then 

P = I (IR) 
but  IR  = E cos  0 
P = El  cos  0 

This  cos  0 is  known  as  the  power  factor  of 
the  circuit.  In  many  circuits  we  strive  to  keep 
the  angle  0 as  small  as  possible,  making  cos  0 
as  near  to  unity  as  possible.  In  tuned  circuits, 
we  use  reactances  which  should  have  as  low  a 
power  factor  as  possible.  The  merit  of  a coil 
or  condenser,  its  Q,  is  defined  by  the  tangent  of 
this  phase  angle: 

Q = ton  « = X/R 

For  an  efficient  coil  or  condenser,  Q should 
be  as  large  as  possible;  the  phase-angle  should 
then  be  as  close  to  90  degrees  as  possible,  mak- 
ing the  power  factor  nearly  zero.  Q is  almost 
but  not  quite  the  inverse  of  cos  0.  Note  that  in 
Figure  29 

Q =X/R  and  cos  0 = R/Z 

When  Q is  more  than  5,  the  power  factor  is 
less  than  20%;  we  can  then  safely  say  Q = 
1/cos  0 with  a maximum  error  of  about  2^ 
percent,  for  in  the  worst  case,  when  cos  0 — 
0.2,  Q will  equal  tan  0 = 4.89.  For  higher 
values  of  Q,  the  error  becomes  less. 

Note  that  from  Figure  29  can  be  seen  the 
simple  relation: 

Z = R -f  jXi. 

|z|  = V R*  + XL' 


Graphical  Representation  773 


Figure  28. 

VECTOR  REPRESENTATION  OF  A 
SIMPLE  SERIES  CIRCUIT. 


The  righihand  portion  of  the  itiusiration  shows 
the  vectors  representing  the  voiiage  drops  in 
the  coil  and  resistance  illustrated  at  the  left. 
Note  that  the  voltage  drop  across  the  coil  Xl 
leads  that  across  the  resistance  by  90 

Graphical  Representation 

Formulas  and  physical  laws  are  often  pre- 
sented in  graphical  form;  this  gives  us  a 
"bird's  eye  view”  of  various  possible  conditions 
due  to  the  variations  of  the  quantities  involved. 
In  some  cases  graphs  permit  us  to  solve  equa- 
tions with  greater  ease  than  ordinary  algebra. 

Coordinate  Systems  All  of  us  have  used  co- 
ordinate systems  with- 
out realizing  it.  For  instance,  in  modern  cities 
we  have  numbered  streets  and  numbered  ave- 
nues. By  this  means  we  can  define  the  location 
of  any  spot  in  the  city  if  the  nearest  street 
crossings  are  named.  This  is  nothing  but  an 
application  of  Cartesian  coordinates. 

In  the  Cartesian  coordinate  system  (named 
after  Descartes),  we  define  the  location  of  any 
point  in  a plane  by  giving  its  distance  from 
each  of  two  perpendicular  lines  or  axes.  Figure 
30  illustrates  this  idea.  The  vertical  axis  is 
called  the  Y-axis,  the  horizontal  axis  is  the 
X-axis.  The  intersection  of  these  two  axes  is 
called  the  origin,  0.  The  location  of  a point, 
P,  (Figure  30)  is  defined  by  measuring  the 
respective  distances,  x and  j along  the  X-axis 
and  the  Y-axis.  In  this  example  the  distance 
along  the  X-axis  is  2 units  and  along  the  Y- 
axis  is  3 units.  Thus  we  define  the  point  as 


R 


Figure  29. 

The  figure  of  merit  of  a coil  ami  its  resistartce 
Is  represented  by  the  ratio  of  the  inductive 
reactance  to  the  resistance,  which  as  shown 
X 

In  this  diagram  is  equal  to  -—  which  equals 

tan  0.  For  large  values  of  0 (the  phase  angle) 
this  IS  approximately  equal  to  the  reciprocal 
of  the  cos  8. 


774  Radio  Mathematics  and  Calculations  THE  RADIO 


Y 


~i 

SE 

QU 

1 

CO 

ADR 

ND 

AN 



1 

FIR 

UAD 

1 

ST 

RA^ 

r 

1 

0 

•IT 

“1 

n 

1 

p 

1 

_j 

9 

8- 

7 - 

6- 

5 - 

4 - 

3 - 

2 - 



0 

1 

\ 

> 

b 

8 

r 

-2 

-3 

-4 

-5 

-6 

-7 

-fl 

F 

r 

r 

1 

TH 

RD 
RAh 
1 

=OL 

UAC 

1 

RTh 
RA 
I 

n 

QUAD 

/T 

Q 

1 

4T 



1 

L_ 

Y 


Figure  30. 

CARTESIAN  COORDINATES. 

The  location  oi  any  point  can  be  defined  by 
its  distance  from  the  X and  Y axes. 


P 2,  3 or  we  might  say  x = 2 and  y — i.  The 
measurement  x is  called  the  abscissa  of  the 
point  and  the  distance  y is  called  its  ordinate. 
It  is  arbitrarily  agreed  that  distances  measured 
from  0 to  the  ri^t  along  the  X-axis  shall  be 
reckoned  positive  and  to  the  left  negative.  Dis- 
tances measured  along  the  T-axis  are  positive 
when  measured  upwards  from  0 and  negative 
when  measured  downwards  from  0.  This  is 
illustrated  in  Figure  30.  The  two  axes  divide 
the  plane  area  into  four  parts  called  quadrants. 
These  four  quadrants  are  numbered  as  shown 
in  the  figure. 

It  follows  from  the  foregoing  statements, 
that  points  lying  within  the  first  quadrant  have 
both  X and  y positive,  as  is  the  case  with  the 
point  P.  A point  in  the  second  quadrant  has  a 
negative  abscissa,  x,  and  a positive  ordinate,  y. 
This  is  illustrated  by  the  point  Q,  which  has 
the  coordinates  x = — 4 and  y = -f  1.  Points 
in  the  third  quadrant  have  both  x and  y nega- 
tive. X = — 5 and  y = — 2 illustrates  such  a 
point,  i?.  The  point  S,  in  the  fourth  quadrant 
has  a negative  ordinate,  y and  a positive  ab- 
scissa or  X. 

In  practical  applications  we  might  draw 
only  as  much  of  this  plane  as  needed  to  illus- 
trate our  equation  and  therefore,  the  scales 
along  the  X-axis  and  Y-axis  might  not  start 
with  zero  and  may  show  only  that  part  of  the 
scale  which  interests  us. 

Representotion  of  In  the  equation: 

Functions 

300,000 


Figure  31. 

REPRESENTATION  OF  A SIMPLE 
FUNCTION  IN  CARTESIAN  CO- 
ORDINATES. 

300,000 

In  this  chart  of  the  function  fke  = x 

K meters 

distances  along  the  X axis  represent  wave- 
length  in  meters,  while  those  along  the  Y 
axis  represent  frequency  in  kilocycles.  A curve 
such  as  this  helps  to  find  vaiues  between 
those  calculated  with  sufficient  accuracy  for 
most  purposes. 


f is  said  to  be  a function  of  X.  For  every  value 
of  / there  is  a definite  value  of  X.  A variable  is 
said  to  be  a function  of  another  variable  when 
for  every  possible  value  of  the  latter,  or  inde- 
pendent variable,  there  is  a definite  value  of 
the  first  or  dependent  variable.  For  instance, 
if  y = 5x^  y is  a function  of  x and  x is  called 
the  independent  variable.  When  a = 3b’  -f  5b’ 
— 25b  ■+•  6 then  a is  a function  of  i>. 

A function  can  be  illustrated  in  our  coordi- 
nate system  as  follows.  Let  us  take  the  equa- 
tion for  frequency  versus  wavelength  as  an 
example.  Given  different  values  to  the  inde- 
pendent variable  find  the  corresponding  values 
of  the  dependent  variable.  Then  plot  the  points 
represented  by  the  different  sets  of  two  values. 


f.c 

XiiiBt  e 

600 

500 

800 

375 

1000 

300 

1200 

250 

1400 

214 

1600 

187 

1800 

167 

2000 

150 

Plotting  these  points  in  Figure  31  and  draw- 
ing a smooth  curve  through  them  gives  us  the 
curve  or  graph  of  the  equation.  This  curve 
will  help  us  find  values  of  / for  other  values 
of  X (those  in  between  the  points  calculated) 
and  so  a curve  of  an  often-used  equation  may 
serve  better  than  a table  which  always  has 
gaps. 

When  using  the  coordinate  system  described 
so  far  and  when  measuring  linearly  along  both 
axes,  there  are  some  definite  rules  regarding 


HANDBOOK 


Representation  of  Functions  775 


Only  two  points  are  needed  to  define  func- 
tions which  resuit  in  a straight  line  as  shown 
in  this  diagram  representing  Ohm's  Law. 


the  kind  of  curve  we  get  for  any  type  of 
equation.  In  fact,  an  expert  can  draw  the  curve 
with  but  a very  few  plotted  points  since  the 
equation  has  told  him  what  kind  of  curve  to 
expect. 

First,  when  the  equation  can  be  reduced  to 
the  form  y = mx  + b,  where  x and  y are  the 
variables,  it  is  known  as  a linear  or  /irst  degree 
function  and  the  curve  becomes  a straight  line. 
(Mathematicians  still  speak  of  a "curve”  when 
it  has  become  a straight  line.) 

When  the  equation  is  of  the  second  degree, 
that  is,  when  it  contains  terms  like  x‘  ot  f 
or  xy.  the  graph  belongs  to  a group  of  curves, 
called  conic  sections.  These  include  the  circle, 
the  ellipse,  the  parabola  and  the  hyperbola.  In 
the  example  given  above,  our  equation  is  of 
the  form 

xy  = e,  e being  equal  to  300,000 

which  is  a second  degree  equation  and  in  this 
case,  the  graph  is  a hyperbola. 

This  type  of  curve  does  not  lend  itself  read- 
ily for  the  purpose  of  calculation  except  near 
the  middle,  because  at  the  ends  a very  large 
change  in  \ represents  a small  change  in  / and 
vice  versa.  Before  discussing  what  can  be  done 
about  this  let  us  look  at  some  other  types  of 
curves. 

Suppose  we  have  a resistance  of  2 ohms 
and  we  plot  the  function  represented  by  Ohm’s 
Law;  JS  = 21.  Measuring  £ along  the  X-axis 
and  amperes  along  the  T-axis,  we  plot  the 
necessary  points.  Since  this  is  a first  degree 
equation,  of  the  form  y = mx  -f  h (for  £ = 
y,  m — 2 and  I = x and  b — 0)  it  will  be  a 
straight  line  so  we  need  only  two  points  to 
plot  it. 

J_  ^ 

(line  passes  through  origin)  q 0 

5 10 

The  line  is  shown  in  Figure  32.  It  is  seen  to 
be  a straight  line  passing  through  the  origin. 


A TYPICAL  GRID  - VOLTAGE 
PLATE-CURRENT  CHARACTER- 
ISTIC CURVE. 

The  equation  represented  by  such  a curve  Is 
so  complicated  that  we  do  not  use  it.  Data 
for  such  a curve  is  obtained  experimentally, 
and  intermediate  values  can  be  found  with 
sufficient  accuracy  from  the  curve. 


If  the  resistance  were  4 ohms,  we  should  get 
the  equation  E = 41  and  this  also  represents 
a line  which  we  can  plot  in  the  same  figure. 
As  we  see,  this  line  also  passes  through  the 
origin  but  has  a different  slope.  In  this  illus- 
tration the  slope  defines  the  resistance  and  we 
could  make  a protractor  which  would  convert 
the  angle  into  ohms.  This  fact  may  seem  incon- 
sequential now,  but  use  of  this  is  made  in  the 
drawing  of  loadlines  on  tube  curves. 

Figure  33  shows  a typical,  grid-voltage, 
plate-current  static  characteristic  of  a triode. 
The  equation  represented  by  this  curve  is 
rather  complicated  so  that  we  prefer  to  deal 
with  the  curve.  Note  that  this  curve  extends 
through  the  first  and  second  quadrant. 

Families  of  curves.  It  has  been  explained 
that  curves  in  a plane  can  be  made  to  illustrate 
the  relation  between  two  variables  when  one 
of  them  varies  independently.  However,  what 
are  we  going  to  do  when  there  are  three  vari- 
ables and  two  of  them  vary  independently.  It 
is  possible  to  use  three  dimensions  and  three 
axes  but  this  is  not  conveniently  done.  Instead 
of  this  we  may  use  a family  of  curves.  We 
have  already  illustrated  this  partly  with  Ohm’s 
Law.  If  we  wish  to  make  a chart  which  will 
show  the  current  through  any  resistance  with 
any  voltage  applied  across  it,  we  must  take  the 
equation  E = IR,  having  three  variables. 

We  can  now  draw  one  line  representing  a 
resistance  of  1 ohm,  another  line  representing 
2 ohms,  another  representing  3 ohms,  etc.,  or 
as  many  as  we  wish  and  the  size  of  our  paper 
will  allow.  The  whole  set  of  lines  is  then 
applicable  to  any  case  of  Ohm’s  Law  falling 
within  the  range  of  the  chart.  If  any  two  of 
the  three  quantities  are  given,  the  third  can  be 
found. 


776  Radio  Mathematics  and  Calculations  THE  RADIO 


Figure  34. 

A FAMILY  OF  CURVES. 

An  equation  such  as  Ohm's  Law  has  three 
variables,  but  can  be  represented  in  Cartesian 
coordinates  by  a family  of  curves  such  as 
shown  here.  If  any  two  quantities  are  given, 
the  third  can  be  found.  Any  point  in  the 
chart  represents  a definite  .value  each  of  C, 
I,  and  R,  which  will  satisfy  the  equation  of 
Ohm's  Law.  Values  of  R not  situated  on  an  R 
line  can  be  found  by  Interpolation. 


AVERAGE  PLATE  CHARACTERISTICS 
Ef=6.3  V. 


Figure  35. 

"PLATE"  CURVES  FOR  A 
TYPICAL  VACUUM  TUBE. 

In  such  curves  we  have  three  variables,  plate 
voltage,  plate  current,  and  grid  bias.  Each 
point  on  a grid  bias  line  corresponds  to  the 
plate  voltage  and  plate  current  represented 
by  its  position  with  respect  to  the  X and  Y 
axes.  Those  for  other  values  of  grid  bias  may 
be  found  by  Interpolation.  The  leadline  shown 
In  the  lower  left  portion  of  the  chart  is  ex- 
plainod  In  the  text. 


Figure  34  shows  such  a family  of  curves  to 
solve  Ohm’s  Law.  Any  point  in  the  chart  rep- 
resents a definite  value  each  of  E,  I,  and  R 
which  will  satisfy  the  equation.  The  value  of 
R represented  by  a point  that  is  not  situated 
on  an  R line  can  be  found  by  interpolation. 

It  is  even  possible  to  draw  on  the  same  chart 
a second  family  of  curves,  representing  a 
fourth  variable.  But  this  is  not  always  possible, 
for  among  the  four  variables  there  should  be 
no  more  than  two  independent  variables.  In 
our  example  such  a set  of  lines  could  represent 
power  in  watts;  we  have  drawn  only  two  of 
these  but  there  could  of  course  be  as  many  as 
desired.  A single  point  in  the  plane  now  indi- 
cates the  four  values  of  E,  I,  R,  and  P which 
belong  together  and  the  knowledge  of  any 
two  of  them  will  give  us  the  other  two  by 
reference  to  the  chart. 

Another  example  of  a family  of  curves  is 
the  dynamic  transfer  characteristic  or  plate 
family  of  a tube.  Such  a chart  consists  of  sev- 
eral curves  showing  the  relation  between  plate 
voltage,  plate  current,  and  grid  bias  of  a tube. 
Since  we  have  again  three  variables,  we  must 
show  several  curves,  each  curve  for  a fixed 
value  of  one  of  the  variables.  It  is  customary 
to  plot  plate  voltage  along  the  X-axis,  plate 
current  along  the  Y-axis,  and  to  make  different 
curves  for  various  values  of  crid  bias.  Such  a 


set  of  curves  is  illustrated  in  Figure  35.  Each 
point  in  the  plane  is  defined  by  three  values, 
which  belong  together,  plate  voltage,  plate 
current,  and  grid  voltage. 

Now  consider  the  diagram  of  a resistance- 
coupled  amplifier  in  Figure  36.  Starting  with 
the  B-supply  voltage,  we  know  that  whatever 
plate  current  flows  must  pass  through  the 
resistor  and  will  conform  to  Ohm’s  Law.  The 
voltage  drop  across  the  resistor  is  subtracted 
from  the  plate  supply  voltage  and  the  remain- 
der is  the  actual  voltage  at  the  plate,  the  kind 
that  is  plotted  along  the  X-axis  in  Figure  35. 
We  can  now  plot  on  the  plate  family  of  the 


PARTIAL  DIAGRAM  OF  A RESIS 
TANCE  COUPLED  AMPLIFIER. 


Th9  portion  of  th9  supply  voltage  wasted 
across  the  50,000‘Ohm  resistor  Is  represented 
in  Figure  35  as  the  loadline. 


HANDBOOK 


Logarithmic  Scales  777 


tube  the  loadline,  that  is  the  line  showing 
which  part  of  the  plate  supply  voltage  is  across 
the  resistor  and  which  part  across  the  tube  for 
any  value  of  plate  current.  In  our  example,  let 
us  suppose  the  plate  resistor  is  50,000  ohms. 
Then,  if  the  plate  current  were  zero,  the  volt- 
age drop  across  the  resistor  would  he  zero  and 
the  full  plate  supply  voltage  is  across  the  tube. 
Our  first  point  of  the  loadline  is  E — 250, 
I — 0.  Next,  suppose,  the  plate  current  were 
1 ma.,  then  the  voltage  drop  across  the  resistor 
would  be  50  volts,  which  would  leave  for  the 
tube  200  volts.  The  second  point  of  the  load- 
line  is  then  E = 200,  7 = 1.  We  can  continue 
like  this  but  it  is  unnecessary  for  we  shall  find 
that  it  is  a straight  line  and  two  points  are 
sufficient  to  determine  it. 

This  loadline  shows  at  a glance  what  hap- 
pens when  the  grid-bias  is  changed.  Although 
there  are  many  possible  combinations  of  plate 
voltage,  plate  current,  and  grid  bias,  we  are 
now  restricted  to  points  along  this  line  as  long 
as  the  50,000  ohm  plate  resistor  is  in  use.  This 
line  therefore  shows  the  voltage  drop  across 
the  tube  as  well  as  the  voltage  drop  across  the 
load  for  every  value  of  grid  bias.  Therefore,  if 
we  know  how  much  the  grid  bias  varies,  we 
can  calculate  the  amount  of  variation  in  the 
plate  voltage  and  plate  current,  the  amplifi- 
cation, the  power  output,  and  the  distortion. 

Logarithmic  Scales  Sometimes  it  is  conven- 
ient to  measure  along 
the  axes  the  logarithms  of  our  variable  quan- 
tities. Instead  of  actually  calculating  the  logar- 
ithm, special  paper  is  available  with  logarith- 
mic scales,  that  is,  the  distances  measured 
along  the  axes  are  proportional  to  the  logar- 
ithms of  the  numbers  marked  on  them  rather 
than  to  the  numbers  themselves. 

There  is  semi-logarithmic  paper,  having 
logarithmic  scales  along  one  axis  only,  the 
other  scale  being  linear.  We  also  have  full 
logarithmic  paper  where  both  axes  carry  log- 
arithmic scales.  Many  curves  are  greatly  sim- 
plified and  some  become  straight  lines  when 
plotted  on  this  paper. 

As  an  example  let  us  take  the  wavelength- 
frequency  relation,  charted  before  on  straight 
cross-section  paper. 

f _ 300,000 
X 

Taking  logarithms: 

lag  f = lag  300,000  — lag  X 

If  we  plot  log  / along  the  Y-axis  and  log  X 
along  the  X-axis,  the  curve  becomes  a straight 
line.  Figure  37  illustrates  this  graph  on  full 
logarithmic  paper.  The  graph  may  be  read 
with  the  same  accuracy  at  any  point  in  con- 


Many  functions  become  greatly  simplified  and 
some  Jbecome  straight  lines  when  plotted  to 
logarithmic  scales  such  as  shown  in  this 
diagram.  Here  the  frequency  versus  wavelength 
curve  of  Figure  3 7 has  been  replotted  to  con- 
form with  logarithmic  axes.  Note  that  it  is 
only  necessary  to  calculate  two  points  in 
order  to  determine  the  *‘curve**  since  this  type 
of  function  results  in  a straight  line. 


trast  to  the  graph  made  with  linear  coordi- 
nates. 

This  last  fact  is  a great  advantage  of  logar- 
ithmic scales  in  general.  It  should  be  clear  that 
if  we  have  a linear  scale  with  100  small  divi- 
sions numbered  from  1 to  100,  and  if  we  are 
able  to  read  to  one  tenth  of  a division,  the 
possible  error  we  can  make  near  100,  way  up 
the  scale,  is  only  1/lOth  of  a percent.  But  near 
the  beginning  of  the  scale,  near  1,  one  tenth  of 
a division  amounts  to  10  percent  of  1 and  we 
are  making  a 10  percent  error. 

In  any  logarithmic  scale,  our  possible  error 
in  measurement  or  reading  might  be,  say  1/32 
of  an  inch  which  represents  a fixed  amount  of 
the  log  depending  on  the  scale  used.  The  net 
result  of  adding  to  the  logarithm  a fixed  quan- 
tity, as  0.01,  is  that  the  anti-logarithm  is  mul- 
tiplied by  1,025,  or  the  error  is  21/2%-  No  mat- 
ter at  what  part  of  the  scale  the  0.01  is  added, 
the  error  is  always  2V2%- 

An  example  of  the  advantage  due  to  the  use 


778  Radio  Mathematics  and  Calculations 


Fi9ure  38. 

A RECEIVER  RESONANCE  CURVE. 

This  curve  represents  the  output  of  a re- 
ceiver versus  frequency  when  plotted  to  linear 
coordinates. 


of  semi-logarithmic  paper  is  shown  in  Figures 
38  and  39.  A resonance  curve,  when  plotted  on 
linear  coordinate  paper  will  look  like  the  curve 
in  Figure  38.  Here  we  have  plotted  the  output 
of  a receiver  against  frequency  while  the  ap- 
plied voltage  is  kept  constant.  It  is  the  kind  of 
curve  a "wobbulator”  will  show.  The  curve 
does  not  give  enough  information  in  this  form 
for  one  might  think  that  a signal  10  kc.  off 
resonance  would  not  cause  any  current  at  all 
and  is  tuned  out.  However,  we  frequently  have 
off  resonance  signals  which  are  1000  times  as 
strong  as  the  desired  signal  and  one  cannot 
read  on  the  graph  of  Figure  38  how  much  any 
signal  is  attenuated  if  it  is  reduced  more  than 
about  20  times. 

In  comparison  look  at  the  curve  of  Figure 
39.  Here  the  response  (the  current)  is  plotted 
in  logarithmic  proportion,  which  allows  us  to 
plot  clearly  how  far  off  resonance  a signal  has 
to  be  to  be  reduced  100,  1,000,  or  even  10,000 
times. 

Note  that  this  curve  is  now  "upside  down”; 
it  is  therefore  called  a selectivity  curve.  The 
reason  that  it  appears  upside  down  is  that  the 
method  of  measurement  is  different.  In  a se- 
lectivity curve  we  plot  the  increase  in  signal 
voltage  necessary  to  cause  a standard  output 
off  resonance.  It  is  also  possible  to  plot  this  in- 
crease along  the  T-axis  in  decibels;  the  curve 
then  looks  the  same  although  linear  paper  can 


Figure  39. 

A RECEIVER  SELECTIVITY  CURVE. 

This  curve  represents  the  selectivity  of  a re- 
ceiver plotted  to  logarithmic  coordinates  for 
the  output,  but  linear  coordinates  for  fre- 
quency. The  reason  that  this  curve  appears 
inverted  from  that  of  Figure  38  is  explained 
in  the  text. 


be  used  because  now  our  unit  is  logarithmic. 

An  example  of  full  logarithmic  paper  being 
used  for  families  of  curves  is  showm  in  the  re- 
actance charts  of  Figures  40  and  41. 

Nomograms  or  An  alignment  chart  con- 

Alignment  Charts  sists  of  three  or  more  sets 
of  scales  which  have  been 
so  laid  out  that  to  solve  the  formula  for  which 
the  chart  was  made,  we  have  but  to  lay  a 
straight  edge  along  the  two  given  values  on 
any  two  of  the  scales,  to  find  the  third  and 
unknown  value  on  the  third  scale.  In  its  sim- 


yiOOJUHY  _\nOJUHY 


Figure  40. 

TANCE-FREQUENCY  CHART 

See  text  for  applicationt  and  ii 

e ..  ^ 


I ji(Si 

l>»liaiKSfi^r<KKS9S> 

I I ■ ■ 9 K*  I 

~ ■ B M « h.'M  kr^»:«.*v 


:4lk!t^9FT4P^.mP^^<iK| 

n9K9:^Ki6^^K^^y.- 

I ■ B K » «K  IK*:-  rw'J'rm 
I n V dk  « riB  »'« »»:•  fc'vy^M 


S»!linA»i>:cr.l!^5»r.F',li!«K«AST<Ci^x^^QPS»U:^IK!«KI^WVT«^'V‘^P«VVV^IfiKPS.%>r<»V<^MSff^T<^J 

mrd. ^ 

iiBS<aA^>r<i;«T-«:^(SFS«K<ia^Bn3K.^^iv^o&B>:^iB!iKF&^K.«^r4:r^pweV'eik!«K«si^K<i^>'r«:^wv:r^| 
1 1 ^ Bii  hr4i  »>:♦:  K’wr^m  c a b w »:«  ».♦:«*  : a a a:  a ‘..a  w«'*a  I 

■■•rjuia.'ri ■■  ~ -_  _ 


mmtmisst 

IBB  K K »; 


I k « K 1^  rM^:f/.  I 

iikn  I 

■a  fc.Tdi?  ►.♦  V 


i I ■ K ■ V ar  A' 'evw*  I B K « ar  #^#srA  I ■ a a:  a AX'*  Mvxe?  I a » « *.'«  Ik' wa* «» ' r.ar/x^^  I 

? ■ ■ ^ ^ ? a ■ u ■ ^ k a a ^ ^ ^%FX.4a ; a a V a •'’.d  •.♦ : V xT^/^a 


lOOOJHY 


1.0JUHY 


0.1  JUHY 


I i>:mm  wk«  | 


it » Ji  Hik  V4  .w  mr^  .'*.*:«si  aw:mm  V4  :%*:  •»:«<  ■ k ■ ^ V4  ■»■■ 




vi»T*T4K4Ki*^^^iiw  aBBPSik:«  :«>:«:4K»r«':tf^w  saHBB." 

J»Jii^>T>«J^i>^:<«JBMB^  gBfSBIP>>T«^<«»^If;<J^M^Ki  gBlSB»t:»:iWlWOu^.<i^B ._ 

I !iiK:4Si«>:^»!<«i»i«  w:s^yL»yjX»z^  »s9Siaisit;!T«i«fiiK<Kiiis9a(!K^^r:;^ 

5i 


>m%<!H^a^s^2WBK£is^SMiai^sisgsi!%sn^»;s»si 


I wr4m'M'^iKrAir^msmsrm  i^si»  WM  i^»v»  ■►TiMK^nfS  w«iw«  l►:«Ri0>:^v:•v'^« 

liiaeiB»il8eBra«SB«»SgBHii8gBRIil8eBi^l 


100  KC 

1000  000  A 


100000A 


10  000 A 


100  KC. 
1 A 


Figure  41. 

REACTANCE-FREQUENCY  CHART  FOR  R.F. 

This  chart  is  used  in  conjunction  with  the  nomogroph  on  poge  569  for  radio  frequency  tank  coil  computations. 


780  Rad  io  Mathematics  and  Calculations 


HANDBOOK 


Polar  Coordinates  781 


plest  form,  it  is  somewhat  like  the  lines  in  Fig- 
ure 42.  If  the  lines  a,  b,  and  c are  parallel  and 
equidistant,  we  know  from  ordinary  geometry, 
that  b = 1/2  (a  + c).  Therefore,  if  we  draw  a 
scale  of  the  same  units  on  all  three  lines,  start- 
ing with  zero  at  the  bottom,  we  know  that  by 
laying  a straight-edge  across  the  chart  at  any 
place,  it  will  connect  values  of  a,  b,  and  c, 
which  satisfy  the  above  equation.  When  any 
two  quantities  are  known,  the  third  can  be 
found. 

If,  in  the  same  configuration  we  used  loga- 
rithmic scales  instead  of  linear  scales,  the  rela- 
tion of  the  quantities  would  become 

log  b = 1/2  (log  a + log  c)  or  b = Vac 

By  using  different  kinds  of  scales,  different 
units,  and  different  spacings  between  the  scales, 
charts  can  be  made  to  solve  many  kinds  of 
equations. 

If  there  are  more  than  three  variables  it  is 
generally  necessary  to  make  a double  chart, 
that  is,  to  make  the  result  from  the  first  chart 
serve  as  the  given  quantity  of  the  second  one. 
Such  an  example  is  the  chart  for  the  design  of 
coils  illustrated  in  Figure  45.  This  nomogram 
is  used  to  convert  the  inductance  in  microhen- 
ries to  physical  dimensions  of  the  coil  and  vice 
versa.  A pin  and  a straight  edge  are  required. 
The  method  is  shown  under  "R.  F.  Tank  Cir- 
cuit Calculations”  later  in  this  chapter. 

Polor  Coordinates  Instead  of  the  Cartesian 
coordinate  system  there 
is  also  another  system  for  defining  algebraical- 
ly the  location  of  a point  or  line  in  a plane.  In 
this,  the  polar  coordinate  system,  a point  is  de- 
termined by  its  distance  from  the  origin,  O, 
and  by  the  angle  it  makes  with  the  axis  0-X, 
In  Figure  43  the  point  P is  defined  by  the 
length  of  OP,  known  as  the  radius  vector  and 


✓ 


/ 


RADIUS  / 
VECTOR  / 

/ 


/ v/tC. 


ORIGIN 


AXIS 


Figure  43. 

THE  LOCATION  OF  A POINT  BY 
POLAR  COORDINATES. 

In  the  polar  coordinate  system  any  point  is 
determined  by  its  distance  from  the  origin 
and  the  angle  formed  by  a fine  drawn  from 
it  to  the  origin  and  the  0~X  axis. 


by  the  angle  A the  vectorial  angle.  We  give 
these  data  in  the  following  form 

P = 3 ^60° 

Polar  coordinates  are  used  in  radio  chiefly 
for  the  plotting  of  directional  properties  of  mi- 
crophones and  antennas.  A typical  example  of 
such  a directional  characteristic  is  shown  in 
Figure  44.  The  radiation  of  the  antenna  rep- 
resented here  is  proportional  to  the  distance  of 
the  characteristic  from  the  origin  for  every 
possible  direction. 


N 


THE  RADIATION  CURVE  OF  AN 
ANTENNA. 

Polar  coordinates  are  used  principally  In  radio 
work  for  plotting  the  directional  characteiis- 
tics  of  an  antenna  where  the  radiation  is 
represented  by  the  distance  of  the  curve  from 
the  origin  for  every  possible  direction. 


782  Rad  io  Mathematics  and  Calculations 


Reactance  Calculations 

In  audio  frequency  calculations,  an  accuracy 
to  better  than  a few  per  cent  is  seldom  re- 
quired, and  when  dealing  with  calculations  in- 
volving inductance,  capacitance,  resonant  fre- 
quency, etc.,  it  is  much  simpler  to  make  use  of 
reactance-frequency  charts  such  as  those  in 
figures  40  and  41  rather  than  to  wrestle  with  a 
combination  of  unwieldy  formulas.  From  these 
charts  it  is  possible  to  determine  the  reactance  of 
a condenser  or  coil  if  the  capacitance  or  induc- 
tance is  known,  and  vice  versa.  It  follows  from 
this  that  resonance  calculations  can  be  made 
directly  from  the  chart,  because  resonance 
simply  means  that  the  inductive  and  capacitive 
reactances  are  equal.  The  capacity  required  to 
resonate  with  a given  inductance,  or  the  induc- 
tance required  to  resonate  with  a given  capac- 
ity, can  be  taken  directly  from  the  chart. 

While  the  chart  may  look  somewhat  formid- 
able to  one  not  familiar  with  charts  of  this 
type,  its  application  is  really  quite  simple,  and 
can  be  learned  in  a short  while.  The  following 
example  should  clarify  its  interpretation. 

For  instance,  following  the  lines  to  their  in- 
tersection, we  see  that  0.1  hy.  and  0.1  /tfd.  in- 
tersect at  approximately  1,500  cycles  and  1,000 
ohms.  Thus,  the  reactance  of  either  the  coil  or 
condenser  taken  alone  is  about  1000  ohms,  and 
the  resonant  frequency  about  1,500  cycles. 

To  find  the  reactance  of  0.1  hy.  at,  say, 
10,000  cycles,  simply  follow  the  inductance 
line  diagonally  up  towards  the  upper  left  till  it 
intersects  the  horizontal  10,000  kc.  line.  Fol- 
lowing vertically  downward  from  the  point  of 
intersection,  we  see  that  the  reactance  at  this 
frequency  is  about  6000  ohms. 

To  facilitate  use  of  the  chart  and  to  avoid 
errors,  simply  keep  the  following  in  mind:  The 
vertical  lines  indicate  reactance  in  ohms,  the 
horizontal  lines  always  indicate  the  frequency, 
the  diagonal  lines  sloping  to  the  lower  right 
represent  inductance,  and  the  diagonal  lines 
sloping  toward  the  lower  left  indicate  capaci- 
tance. Also  remember  that  the  scale  is  loga- 
rithmic. For  instance,  the  next  horizontal  line 
above  1000  cycles  is  2000  cycles.  Note  that 
there  are  9,  not  10,  divisions  between  the  heavy 
lines.  This  also  should  be  kept  in  mind  when 
interpolating  between  lines  when  best  possible 
accuracy  is  desired;  halfway  between  the  line 
representing  200  cycles  and  the  line  represent- 
ing 300  cycles  is  not  250  cycles,  but  approxi- 
mately 230  cycles.  The  250  cycle  point  is  ap- 
proximately 0.7  of  the  way  between  the  200 
cycle  line  and  the  300  cycle  line,  rather  than 
halfway  between. 

Use  of  the  chart  need  not  be  limited  by  the 
physical  boundaries  of  the  chart.  For  instance, 
the  10-/i/ifd.  line  can  be  extended  to  find  where 


it  intersects  the  100-hy.  line,  the  resonant  fre- 
quency being  determined  by  projecting  the  In- 
tersection horizontally  back  on  to  the  chart. 
To  determine  the  reactance,  the  logarithmic 
ohms  scale  must  be  extended. 

R.  F.  Tonk  When  winding  coils  for  use  in 
Circuit  radio  receivers  and  transmit- 

Calculotions  ters,  it  is  desirable  to  be  able  to 
determine  in  advance  the  full 
coil  specifications  for  a given  frequency.  Like- 
wise, it  often  is  desired  to  determine  how  much 
capacity  is  required  to  resonate  a given  coil  so 
that  a suitable  condenser  can  be  used. 

Fortunately,  extreme  accuracy  is  not  re- 
quired, except  where  fixed  capacitors  are  used 
across  the  tank  coil  with  no  provision  for  trim- 
ming the  tank  to  resonance.  Thus,  even  though 
it  may  be  necessary  to  estimate  the  stray  cir- 
cuit capacity  present  in  shunt  with  the  tank 
capacity,  and  to  take  for  granted  the  likelihood 
of  a small  error  when  using  a chart  instead  of 
the  formula  upon  which  the  chart  was  based, 
the  results  will  be  sufficiently  accurate  in  most 
cases,  and  in  any  case  give  a reasonably  close 
point  from  which  to  start  "pruning.” 

The  inductance  required  to  resonate  with  a 
certain  capacitance  is  given  in  the  chart  in 
figure  41.  By  means  of  the  r.f.  chart  , the 
inductance  of  the  coil  can  be  determined, 
or  the  capacitance  determined  if  the  induc- 
tance is  known.  When  making  calculations,  be 
sure  to  allow  for  stray  circuit  capacity,  such  as 
tube  interelectrode  capacity,  wiring,  sockets, 
etc.  This  will  normally  run  from  5 to  25  micro- 
microfarads, depending  upon  the  components 
and  circuit. 

To  convert  the  inductance  in  microhenries 
to  physical  dimensions  of  the  coil,  or  vice 
versa,  the  nomograph  chart  in  figure  41  is 
used.  A pin  and  a straightedge  ate  required. 
The  inductance  of  a coil  is  found  as  follows: 

The  straightedge  is  placed  from  the  correct 
point  on  the  turns  column  to  the  correct  point 
on  the  diameter-to-length  ratio  column,  the 
latter  simply  being  the  diameter  divided  by  the 
length.  Place  the  pin  at  the  point  on  the  plot 
axis  column  where  the  straightedge  crosses  it. 
From  this  point  lay  the  straightedge  to  the  cor- 
rect point  on  the  diameter  column.  The  point 
where  the  straightedge  intersects  the  induc- 
tance column  will  give  the  inductance  of  the 
coil. 

From  the  chart,  we  see  that  a 30  turn  coil 
having  a diameter-to-length  ratio  of  0.7  and  a 
diameter  of  1 inch  has  an  inductance  of  ap- 
proximately 12  microhenries.  Likewise  any  one 
of  the  four  factors  may  be  determined  if  the 
other  three  are  known.  For  instance,  to  deter- 
mine the  number  of  turns  when  the  desired  in- 


N»  OF 
TURNS 

— 400 

— 300 

— 200 

— 150 

— 100 

— 90 

— ao 

— 70 

— 60 
— SO 

— 40 

- r-ao" 

~20 

— 15 

— 10 

— 5 

— 4 

— 3 


Figure  45.  COIL  CALCULATOR  NOMOGRAPH 

For  single  layer  solenoid  coils,  any  wire  size.  See  text  for  instructions. 


PLOT  INDUCTANCE  IN 
AXIS  MICROHENRIES 


— 20000 

— 10000 
^8000 
— 6000 

— 4000 

— 3000 

— 2000 

— 1000 
^800 
— 600 

— 400 

— 300 

— 200 


RATIO- 


DIAMETER 

LENGTH 

8 —t 


8- 

5- 

4- 


2H 


t-f 


DIAMETER 
INCHES 
r— S 


^-4 


H3 


I—  2 


784  Radio  AAathematics  and  Calculations 


ductance,  the  D/L  ratio,  and  the  diameter  are 
known,  simply  work  backwards  from  the  ex- 
ample given.  In  all  cases,  remember  that  the 
straightedge  reads  either  turns  and  D/L  ratio, 
or  it  reads  inductance  and  diameter.  It  can 
read  no  other  combination. 

The  actual  wire  size  has  negligible  effect 
upon  the  calculations  for  commonly  used  w'ire 
sizes  (no.  10  to  no.  30).  The  number  of  turns 
of  insulated  wire  that  can  be  wound  per  inch 
(solid)  will  be  found  in  a copper  wire  table. 

Significant  Figures 

In  most  radio  calculations,  numbers  repre- 
sent quantities  which  were  obtained  by  meas- 
urement. Since  no  measurement  gives  absolute 
accuracy,  such  quantities  are  only  approximate 
and  their  value  is  given  only  to  a few  signifi- 
cant figures.  In  calculations,  these  limitations 
must  be  kept  in  mind  and  one  should  not  fin- 
ish for  instance  with  a result  expressed  in  more 
significant  figures  than  the  given  quantities  at 
the  beginning.  This  would  imply  a greater  ac- 
curacy than  actually  was  obtained  and  is  there- 
fore misleading,  if  not  ridiculous. 

An  example  may  make  this  clear.  Many  am- 
meters and  voltmeters  do  not  give  results  to 
closer  than  V4  ampere  or  Va  volt.  Thus  if  we 
have  2V4  amperes  flowing  in  a d.c.  circuit  at 
6^/4  volts,  we  can  obtain  a theoretical  answer 
by  multiplying  2.25  by  6.75  to  get  15.1875 
watts.  But  it  is  misleading  to  express  the  an- 
swer down  to  a ten-thousandth  of  a watt  when 
the  original  measurements  were  only  good  to 
1/4  ampere  or  volt.  The  answer  should  be  ex- 
pressed as  15  watts,  not  even  15.0  watts.  If 
we  assume  a possible  error  of  Vs  volt  or  am- 
pere (that  is,  that  our  original  data  are  only 
correct  to  the  nearest  Vi*  volt  or  ampere)  the 
true  power  lies  between  14.078  (product  of 
21/8  and  65/g)  and  16.328  (product  of  and 
6%).  Therefore,  any  third  significant  figure 
would  be  misleading  as  implying  an  accuracy 
which  we  do  not  have. 

Conversely,  there  is  also  no  point  to  calcu- 
lating the  value  of  a part  down  to  5 or  6 sig- 
nificant figures  when  the  actual  part  to  be  used 
cannot  be  measured  to  better  than  1 part  in 
one  hundred.  For  instance,  if  we  are  going  to 
use  1%  resistors  in  some  circuit,  such  as  an 
ohmmeter,  there  is  no  need  to  calculate  the 
value  of  such  a resistor  to  5 places,  such  as 
1262.5  ohm.  Obviously,  1%  of  this  quantity 
is  over  12  ohms  and  the  value  should  simply 
be  written  as  1260  ohms. 

There  is  a definite  technique  in  handling 
these  approximate  figures.  When  giving  values 
obtained  by  measurement,  no  more  figures  are 


given  than  the  accuracy  of  the  measurement 
permits.  Thus,  if  the  measurement  is  good  to 
two  places,  we  woqld  write,  for  instance,  6.9 
which  would  mean  that  the  true  value  is 
somewhere  between  6.85  and  6.95.  If  the  meas- 
urement is  known  to  three  significant  figures, 
we  might  write  6.90  which  means  that  the  true 
value  is  somewhere  between  6.895  and  6.905. 
In  dealing  with  approximate  quantities,  the 
added  cipher  at  the  right  of  the  decimal  point 
has  a meaning. 

There  is  unfortunately  no  standardized  sys- 
tem of  writing  approximate  figures  with  many 
ciphers  to  the  left  of  the  decimal  point.  69000 
does  not  necessarily  mean  that  the  quantity  is 
known  to  5 significant  figures.  Some  indicate 
the  accuracy  by  writing  69  x 10’  or  690  x 10’ 
etc.,  but  this  system  is  not  universally  em- 
ployed. The  reader  can  use  his  own  system,  but 
whatever  notation  is  used,  the  number  of  sig- 
nificant figures  should  be  kept  in  mind. 

Working  with  approximate  figures,  one  may 
obtain  an  idea  of  the  influence  of  the  doubtful 
figures  by  marking  all  of  them,  and  products 
or  sums  derived  from  them.  In  the  following 
example,  the  doubtful  figures  have  been  under- 
lined. 

603 

34.6 

O.T20 

637.7M  onswer:  638 

Multiplication. 

654  654 

0.342  0.342 

1308  19612 

2616  26|16 

1962“  11308 

223.668  answer:  224  224 

It  is  recommended  that  the  system  at  the 
right  be  used  and  that  the  figures  to  the  right 
of  the  vertical  line  be  omitted  or  guessed  so  as 
to  save  labor.  Here  the  partial  products  are 
w'ritten  in  the  reverse  order,  the  most  impor- 
tant ones  first. 

In  division,  labor  can  be  saved  when  after 
each  digit  of  the  quotient  is  obtained,  one  fig- 
ure of  the  divisor  be  dropped.  Example: 

1.28 

527  r673 
527 
53  Fh? 

106 

5 

40 


UOW-PASS  T SeCTIOM  UOW-^ASS  TT  SECTION  HICH-PASS  T SECTION  HIGH  - PA  SS  77"  SEC  T ION 


FREQUENCY  SCALE 
LOW- PASS  HIGH- PASS 


VALUES 

L C 


L C 


FILTER  DESIGN  CHART 
For  both  Pi-type  ond  T-type  Sections 

To  find  L,  connect  cut-off  frequency  on  teft-hund  scale  (using  left-side  scale  for  low-pass  and  right- 
side  scale  for  high-pass)  with  load  on  left-hand  side  of  right-hand  scale  by  means  of  a straight-edge. 
Then  read  the  value  of  L from  the  point  where  the  edge  intersects  the  left  side  of  the  center  scale.  Read- 
ings are  in  henries  for  frequencies  in  cycles  per  second. 

To  find  C,  connect  cut-off  frequency  on  left-hand  scale  (usrng  left-side  scale  for  low-pass  and  right- 
side  scale  for  high  pass)  with  the  load  on  the  right-hand  side  of  the  right-hand  scale.  Then  read  the 
value  of  C from  the  point  where  the  straightedge  cuts  the  right  side  of  the  center  scale.  Readings  are 
in  microfarads  for  frequencies  in  cycles  per  second. 

For  frequencies  in  kilocycles,  C is  expressed  in  thousands  of  micromicrofarads,  L is  expressed  in 
millihenries.  Per  frequencies  in  megacycles,  L is  expressed  in  microhenries  and  C is  expressed  in  micromi- 
crofarads. 

For  each  tenfold  increase  In  the  value  of  load  resistance  multiply  L by  10  and  divide  C by  10. 
For  each  tenfold  decrease  In  frequency  multiply  L by  fO  and  multiply  C by  10. 


785 


FOR  USE  WtTH  RIGHT  CENTER  CAPACITANCE  SCALE 


INDEX 


Acceptor,  def.  of 92 

A.C.-D.C.  power  supply 561,  588,  694 

"Acorn”  tube 232 

A-C  V-T  voltmeter 724 

Adapter,  F-M 325 

Adapter,  mechanical  filter 535 

Admittance,  def.  of 52 

"A"-frame  antenna  mast 448 

Air  capacitor 359 

Air  dielectric 32 

Air-gap,  capacitor 264 

Alignment  — B.f.o  — Filter 236 

" Receiver  — l-F  — R-F 148,  235,  236 

All-driven  antenna 504 

Alpha,  def.  of 93 

Alternating  current  — Amplitude  — Average 45 

" Def.  of  — Effective  value 41,  45 

" Generation  42 

" Impedance  — "J"  operator 47 

” Ohm’s  law  — Phase  angle 45,  46 

Pulsating  — R.M.S 45 

" Transformer  61 

" Transient 59 

" Voltage  divider 52 

Alternator  42 

Amateur  band  receiver 540 

Amateur  frequency  bands  — licenses 12 

Amateur  radio  service.  The 11 

Ammeter  721 

Ampere,  def.  of 22,  23 

Ampere-turns  36,  62 

Amperite  filament  regulator 552 

Amplification  factor 73,  74 

Amplifier  — Audio  — Cascode 227,  214 

" Cathode  — Cathode  driven 127,  626 

” Cathode  follower 134 

” Class  A — A2 99,  108,  1 18 

” Class  AB  — ABl 108,  137 

” Class  B 99,  108,  123,  1 29,  251 

” Class  C 108,  124,  251 

" Coupling  — inductive 1 13,  118,  123 

’’  D.C 117 

" Doherty  298 

" Driver  126 

” Equivalent  circuit 109 

" Feedback  129 

" Grid  circuit 1 2 1 

” Grounded  cathode — Grounded  screen  132,629 

” Grounded  grid 132,  135,  214,  258,  626 

■'  Hi-fi  — High  mu  {ix) 138,  112 

" Horizontal  — l-F 139,  210 

” Kilowatt,  general  purpose 635 

’’  Linear  138,  286,  620 

" Loftin-White  1 1 8 

" Multi-band  linear 624 

" Neutralization  252,  616,  639 

" Noise  factor 122 

" Operational  199 

" Pentode  120 

’’  Pi-network  617 

" Plate  efficiency 129 

■’  Power 118,610 

” Printed  circuit 575 


A 

Amplifier  — Push-pull  121,  612 

" Push-pull  tetrode 644 

••  R-F  — RC 121,  109 

" Response  112 

’•  R-F  — 4CX-1000A 213,  640 

” Summing  201 

'*  Terman-Woodyard  298 

" Tetrode  614 

’*  Transistor  104 

" Ultra-linear  142 

" Vertical  138 

*'  V.h.f.  cascode 578 

" Video  112,  113,  128 

*'  Voltage  110 

**  Williamson  140 

Amplitude  — A.C 45 

*'  Distortion  — Modulation 109,  282,  647 

Analog  computers  — problems 195,  200 

Angle  of  radiation  — V.h.f 41  1,  460,  478 

Anode,  def.  of  — dissipation 67,  72 

Antenna  — Adjustment 509 

’’  All-driven  504 

" Bandwidth  405,  413 

” Beam  — design  chart  — 2-element  498,  495 

Bidirectional  ■ — Bi-square  505,  470 

" Bobtail  472 

” Broadside  468 

" Bruce  470 

■’  Center-fed  438 

" Colinear  466 

" Combinations  of  475 

’’  Construction  448,  506 

’’  Corner  reflector  485 

” Couplers  — mobile  454,  520,  524 

” Coupling  systems  451 

” Cubical  quad  471 

" Curtain  468 

" Delta  match  501 

Dipole  array  — broad-band  465,  435 

" Directivity  404,  410 

" Discone  441,  481 

■’  Doublet  427 

” Dummy  726 

” Efficiency  409 

''  Element  spacing  468 

*'  End-effect  406 

’’  End-fed  — end-fire  426,  437,  473 

Feed  systems  428,  498 

" Franklin  array  466 

" Fuchs  426 

” Gain,  beam  495 

*'  Gamma  match  503 

” Ground  loss  409 

" Ground  plane  431 

" Hertz  426 

" High  frequency  type  459 

” Impedance  408 

" Intercept,  v.h.f 477 

" Lazy-H  468 

Length  — Length-to-diameter  ratio  405,  406 

’’  Long  wire  461 

" Maconi  408,  432 

Matching  systems  442 


786 


Anfenna  — Measurements  730 

" Multee  440 

■■  Multi-band  436,  514 

” Mutual  coupling  475 

” Parasitic  beam  494 

" Patterns  404,  460 

" Phasing  466 

" Polarization  404 

" Polarization,  v.h.f 479 

" Power  gain  405 

’’  Q 409 

*'  Reactance  — Resonance  408,  404 

" Rhombic  464 

Rotary  — Rotary  match  494,  502 

” Rotator  512 

" Single  wire  feed  441 

" Six-shooter  472 

" Sleeve  480 

” Space  conserving  434 

" Stacked  468,  498 

" Sterba  469 

” Stub  adjustment  — stub  match  445,  503 

" T-match  501 

" 3-element  beam — 3-element  match  603,  488 

” Triplex  475 

" Tuner  456 

" Turnstile  480 

" Two-band  Marconi  438 

” V-type  462 

" Vertical  430 

V.h.f. — 8-element — ground  plane  ....488,  481 

V.h.f.  — helical  — horn  483,  486 

V.h.f.  nondirectional  482 

V.h.f.  rhombic  486 

V.h.f.  — screen  array  — • Yogi  491,  488 

" W8JK  473 

" X-array  469 

" Yoke  match  498 

" “Zepplin"  427,  467 

Antennascope  738 

Anti-resonance  54 

Aquadag  85 

Area,  capacitor  plate  33 

Arrays,  antenna  465,  412 

“Assumed”  voltage  52 

Atom,  def.  of  — atomic  number  21 

Attenuation,  Bass/Treble  139 

Audio  — Amplifier  227 

" Distortion  135 

” Equalizer  139 

Feedback  loop  141 

“ Filter,  SSB  333 

” Frequency,  def.  of  42 

” Hum  142 

” Limiters  153,  185,  228 

" Phase  Inverter  1 1 5 

" Phasing  network  — SSB  570,  338 

” Wiring  technique  142 

Autodyne  detector 208 

Automatic  load  control,  SSB  345 

Automatic  modulation  control  648 

Automatic  volume  control  225 

Autotransformer  63,  390 

Avalanche  voltage  697 

B 

Balanced  modulator  — SSB  331,  338,  343 

Balanced  SWR  bridge  736 

Balanced  transmission  line  measurements  734 

Bands,  amateur  12 

Bandspread  tuning  217 

Bandwidth,  antenna  405,  413 

Bandwidth,  l-F,  graph  534 


Bandwidth,  modulation  283 

Base  electrode  92 

Bass  suppression,  audio  306 

Battery  bias  269 

Beam  — Design  chart  — Dimensions  498,  496 

” Element  spacing  497 

” Front-back  ratio  496 

” H-F  — 5-element  — 2-element  ....496,  494 

” Power  tube  78 

” Radiation  resistance  495 

” Three  element  HF  — 220  Me 496,  603 

Beat  note  208 

Beat  oscillator  224 

B-H  curve  36 

Bias  — Battery 269 

” Cathode  268 

” Def.  of  — cut-off  73 

" Grid  108,  267 

'*  Grid  leak  112,  268 

” R-F  amplifier  612 

” Safety  268 

” Shift  modulator  307 

Supply,  modulator  653 

” Transistor  96 

Bidirectional  antenna  505 

Bifilar  coil  572 

Billboard  antenna  477 

Binary  notation  195,  196 

Bishop  noise  limiter  228 

Bi-square  antenna  470 

Bistable  multivibrator  102 

Bit  196 

Blanketing  interference  380 

Bleeder  resistor  — safety  27,  709,  395 

Blocked  grid  keying  398 

Blocking  diodes  401 

Blocking  oscillator  102,  155,  189,  190 

Bob-toil  antenna  472 

Bombardment,  cathode  70 

Boost,  Bass/Treble  139 

Break  voltage  203 

Bridge  measurements  728 

Bridge  rectifier  689 

Bridge,  slide-wire  728 

Bridge,  SWR  458,  736 

Bridge-T  oscillator  192 

Bridge-type  vacuum  tube  voltmeter  131 

Bridge,  Wheatstone  728 

Broad  band  dipole  435 

Broadcast  interference  379 

Broadside  antennas  — arrays  468,  412 

Bruce  antenna  470 

Butterfly  circuit  232 

C 

Calorimeter  726 

Copacitance  — Calculation  — definition  ....32,  30 

Interelectrode  76,  106 

Neutralization  107 

Tank  — Stray  261,  218 

Capacitive  coupling  270 

Capacitive  reactance  46 

Capacitor  — A.C.  circuit  34 

Air  — Characteristics  359,  358 

" Breakdown  264 

■■  Charge  31 

” Color  code  531 

” Electrolytic  34^  708 

Equalizing  34 

" Filter  707^  690 

Fixed  

” Leakage  34 

Low  inductance  619 


787 


Capacitor  — Parallel  33 

" Q 54 

Series  33 

” Temperature  coefficient  32 

Time  constant  39 

Vacuum  — variable  vacuum  360,  620 

” Voltage  rating  34 

Carrier  — Distortion  313 

" Shift  284 

" S.S.B 327,  331,  332,  334 

Cascode  amplifier  — V.H.F Ill,  214,  578 

Cathode  — Bias  108,  268 

*’  Coupling  — coupled  inverter  115,  116 

" Current  — Definition  79,  67 

’’  Driven  amplifier  258,  626 

" Follower  274 

” Follower  amplifier 127,  134 

Follower  driver  659,  662 

’’  Follower  modulator  290 

" Keying  397 

" Modulation  297 

Cathode  ray  oscilloscope  — tube  138,  84 

Cavity,  resonant  232 

Cell,  Weston  24 

Center-fed  antenna  438 

Ceramic  dielectric  32 

Charge  22,  23,  30 

Chassis  layout  726 

Choke,  filter  709 

Choke,  R-F  272,  361 

Circuit  constants,  measurement  of  727 

Circuits  — Coupled  123 

D.C.  — Equivalent  21,  51 

” Input  loading  122 

Limiting  151 

" Magnetic  35 

" Parallel  25 

" Parasitic  365 

’’  Q 54 

Resonant  — Series  — Tank  ....53,  24,  55,  57 

Circulating  current  56 

Clamping  circuit  153,  154,  187 

Clapp  oscillator  241 

Class  A amplifier,  def.  of  108 

Class  A2  amplifier  118 

Class  AB  amplifier,  def.  of  108 

Class  ABl  amplifiers  137,  140 

Class  B amplifiers  108,  1 23,  125,  1 29,  139 

Class  B modulator  1 26,  294 

Class  B,  R-F  amplifier  251 

Class  C amplifiers  108,  124,  1 25,  251 

Class  C bias  267 

Clipper-Amplifier  design  657 

Clipper  filter,  speech  304 

Clipping  circuits  151,  185 

Clipping,  negative  peak  648 

Clipping,  phase  shift  301 

Clipping,  speech  124,  300,  648 

Closed  loop  feedback  192 

Coaxial  delay  line  204 

Coaxial  reflectometer  732 

Coaxial  transmission  line  423 

Coaxial  tuned  circuit  232 

Code  14-20 

Code,  practice  set  19 

Coefficient  of  coupling  38 

Coefficient,  temperature  32 

Coercive  force  37 

Coil,  bifilar  572 

Coil,  core  loss  55 

Coil,  Q 54,  360 

Colinear  antenna  466 

Collector  electrode  92 


Collins  feed  system  447 

Collins  filter  222 

Color  code,  standard  531,  532 

Colpitis  oscillator  241 

Complementary  symmetry  97 

Complex  quantity  49 

Component  nomenclature  graph  530 

Compound,  def.  of  21 

Compression,  volume  648 

Compressor,  A.L.C 345 

Computers,  electronic  194 

Conductance,  def.  of 52 

Conduction  67,  90 

Conductivity  — Conductor  23 

Constant  amplitude  recording  137 

Constant  current  curve  74,  1 27 

Constant-K  filter  64 

Constant  velocity  recording  136 

Constants,  vacuum  tube  107 

Construction,  transmitter  — VFO  663,  676 

Control  circuit,  mobile  525 

Control  circuit,  transmitter  391 

Conversion  conductance  80 

Conversion,  frequency,  SSB  340 

Converter  stage  211 

Converter,  transistor  100 

Core  loss  55 

Corner  reflector  antenna  485 

Corona  static  529 

Coulomb  22,  30 

Counter  e.m.f 37 

Counting  circuits  156,  190 

Coupler,  directional  735 

Coupling  — Capactive  270 

’■  Cathode  115 

■*  Choke  114 

'•  Critical  57,  218 

" Direct  115 

" Effect  of  56 

” Impedance  — Inductive  113,  271 

'*  Interstage  — r«f  270,  619 

" Link  272,  61  1 

■'  Magnetic  38 

" Parasitic  364 

••  R-F  feedback  275 

” Systems,  antenna  451 

" Transformer  113 

" Unity  271 

Critical  coupling  218 

Critical  inductance  686 

Crosby  modulotor  590 

Cross  modulation  381 

Crossover  point  136 

Crystal  — Current  246 

" Filler  — Lattice  filter  220,  336,  579 

” Harmonic  246 

" Oscillator  — Quartz  244,  245 

■'  Pickup  137 

Cubical  quad  antenna  471 

Current  — Alternating  41 

” Amplification  (alpha)  93 

Cathode  79 

” Circulating  — Closed  path  56,  28 

’*  Def.  of  22 

" Effective  — Effective  (a.c.)  723,  45 

" Electrode  106 

” Induced  — Instantaneous  42,  44 

’*  Inverse  693 

” Measurement  721,  723 

” Negative  screen  641 

” Peak  amplifier  125 

” Rating,  power  supply  685 

” Saturation  — Skin  effect  72,  54 


788 


Curtain  antenna  468 

Cutoff  (alpha)  93 

Cutoff  bias  73 

Cutoff,  extended  287 

Cutoff  frequency  64 

Cycle  41 

D 

d'Arsonval  meter  721 

D.C.  amplifier  1 1 7 

D.C.  clamping  circuit  187 

D.C.  power  supplies  684 

D.C.  restorer  — circuit  188,  154 

Defector  — autodyne  — crystal  208 

Deflection  plate  139 

Degeneration,  transistor  96 

Degree,  electrical  43 

Delay  line  204 

Delta  match  — antenna  — system  ....501, 430,  442 

Demodulator  207 

Detection,  slope  210 

Detection,  synchronous,  DSB  353 

Detector  — Diode  224 

" Envelope  151 

" Fremodyne  210 

" Gride  leak  — Impedance  — Plate  224 

" Product  237,  352 

**  Ratio  323 

" Super-regenerative  209 

Deviation,  FM  — Measurement  314,  320 

Dielectric,  ceramic  — Constant  (K)  30,  31,  32 

Differential  keying  400 

Differentiation,  electronic  198 

Differentiator  (RC)  59 

Digital  circuits  — computers  195 

Digital  package  204 

Diode  — A.v.c 225 

” Blocking  401 

" Crystal  234 

” Def.  of  71 

" Detector  224 

’’  Gate  204 

” Limiter  185 

Mixer  212 

Modulator  332 

" Semi-conductor  90 

” Storage  time  103 

" Voltmeter  724 

Dipoles  403,  435,  465,  498 

Direct  current  circuits  21 

Directional  coupler  735 

Directive  antennas,  H-F  459 

Directivity,  antenna  404,  410 

Discharge  of  capacitor  30 

Discone  antenna  441,  481 

Discontinuities,  transmission  line  425 

Discriminator,  F-M  322 

Dissipation,  anode  72 

Distortion  — Amplitude  — Frequency  109 

” Audio  135 

" Carrier  313 

” Harmonic  119 

” Intermodulation  136 

" Modulation  307 

” Nonlinear  — Phase  109,  135 

" Products,  SSB  344 

” Transient  135 

” Transistor  98 

Divider,  frequency  190 

Divider,  voltage  26 

Doherty  amplifier  298 

Dome  audio  phasing  network  338 

Double  conversion  circuit  215 


Double  sideband,  DSB  333,  353 

Doubler,  frequency  258 

Doubler,  push-push  260 

Doublet,  antenna  — multi-wire  427,  429 

Drift  transistor  93 

Driver,  amplifier  126 

Driver,  cathode  follower  659,  662 

Duct  propagation  415 

Dummy  antenna  — dummy  loads  726 

Duplex  transmission  597 

Dynamic  resistance  95 

Dynamotor,  PE-103  - 526 

Dynatron  oscillator  242 

E 

Eddy  current  38 

Efficiency,  antenna  409 

Electric  filters  63 

Electrical  energy  29 

Electrical  potential  22 

Electromagnetic  deflection  86 

Electrolytic  capacitor  34,  708 

Electrolytic  conductor  22,  67 

Electromagnetism  35 

Electromotive  force  (e.m.f.)  22,  37 

Electron  coupled  oscillator  241 

Electron,  def.  of  — drift  — orbit  67,  21 

Electronic  computers  194 

Electronic  conduction  67 

Electronic  differentiation  — integration  198 

Electronic  multiplication  198 

Electrostatics  30 

Electrostatic  deflection  — energy  83,  30 

Element,  def.  of  — metallic,  non-metallic  ....21,  90 

Element,  reactive,  non-reactlve  21 

Emmission  equation  70 

Emission,  photoelectric  67 

Emission,  secondary  71,  77 

Emission,  spurious  383 

Emission,  thermionic  67 

Emitter  electrode  92 

Enclosures  724,  725 

End  effect,  antenna  406 

End-fed  antenna  426,  437 

End-fire  antennas  — arrays  473,  412 

Energy  — Electrical  — Electrostatic  29,  30 

” Potential  30 

" Storage,  capacitor  31 

” Transistor  93 

ENIAC  195 

Envelope,  carrier  282 

Envelope  detector  151 

Equalizer,  audio  139 

Equol-tempered  scale  135 

Equivolent  circuit  51 

Equivalent  circuit,  transistor 95 

Equivalent  noise  resistance  123 

Error  cancellation,  feedback  159 

Error  signal  159,  193 

Excitation,  grid  252 

Exciters,  SSB  346,  349 

Extended-Zepp  antenna  467 

F 

Factor  of  merit  54 

Fading  418 

Farad,  def.  of  31 

Feedback  amplifier  129 

Feedback,  audio  141 

Feedback  circuits  192 

Feedback,  control  158 

Feedback,  error  cancellation  159 

Feedback,  Miller  I99 


789 


Feedback,  R-F  274,  615 

Feedback,  transistor  101 

Feed  systems,  antenna  428,  498 

Feed-through  power  627 

Field,  magnetic  35 

Filament,  def.  of  67 

Filament  reactivation  68 

Filament  regulator 552 

Filter  — Capacitor  707,  690 

” Carrier,  SSB  334 

’■  Choke  709 

Circuits,  mobile  518 

’’  Crystal  220 

" Crystal  lattice  — SSB  579,  336 

" Generator,  SSB  333 

” High  pass  64,  372 

" Inductor  input  687 

" Insertion  loss  64 

” Low  pass  64,  376 

M-derived  64 

’’  Mechanical  222,  336 

’’  Noise  227 

" Passband,  SSB  336 

" Q-multiplier  236 

" Resistance-capacitance — Resonance  687,  688 

Ripple  factor  688 

Sections  — Series  — shunt  64 

” TVI,  receiver  — TVI-type  372,  64 

" Wave  63 

Filter-type  exciter,  SSB  349 

Fixed  bias  1 08 

Flat-top  beam  antenna  473 

Fletcher-Munson  curve  140 

Floating  Paraphase  inverter  116 

Flux,  def.  of  — ■ density  35,  36 

Flywheel  effect  57,  259 

Folded  dipole  429,  498 

Forcing  function  200 

Foster-Seely  discriminator  «....322 

Franklin  antenna  466 

Franklin  oscillator  243 

Free  electrons  21 

Free-running  multivibrator  189 

Fremodyne  detector 210 

Frequency  41,  58 

” Conversion,  SSB  340 

■■  Cutoff  64 

" Distortion  109 

” Divider  190 

" Interruption  209 

’’  Maximum  useable  417 

" Measurements  729 

" Multiplier  258 

" Pulse  repetition  190 

” Range,  hi-fidelity  136 

” Ratio,  Lissajous  144 

” Resonant  53 

” Shift  keying  326 

” Sound  134 

" Spectrum  41 

" Spotter  730 

" Standard  729 

” Sweep  1 40 

Frequency  Modulation  312 

” Adapter  325 

” Devotion  — ratio  — index  314 

” Discriminator  322 

" Limiter  324 

" Linearity  318 

” Narrow  band  315 

” Pre-emphasis  325 

" Reception  321 

Front-back  ratio,  beam  496 


Fuchs  antenna  426 

Full-wave  limiter  230 

Full-wave  rectifier  689 

Function  generator  — non-linear  — ramp  202,  203 

G 

G,  conductance,  def.  of  52 

Gain,  antenna  — beam  461,  495 

Gain,  power  — resistance  — voltage  94 

Gamma  match,  antenna  — system  503,  444 

Gas  tube  — generator  87,  140 

Gate,  diode  — limiter  204,  1 85 

Gauss,  def.  of 36 

Generator  noise  5 28,  740 

Generator,  sawtooth  140 

Generator,  time  base  139 

Gilbert,  def.  of  36 

Grid  bias  108,  267 

Grid,  def.  of  72 

Grid  excitation  252 

Grid  leak  bias  109,  1 12,  268 

Grid  leak  detector  2 24 

Grid  limiter,  limiting  153,  187 

Grid  modulation  252,  288 

Grid  neutralization  253 

Grid-screen  mu  factor  78 

Ground  bus  142,  147 

Ground  currents  362 

Ground  loss,  antenna  409 

Ground  plane  antenna  — VHF  431,  481 

Ground  resistance  410 

Ground,  R-F  362 

Ground  termination  433 

Ground,  transmitter  394 

Ground  wave  414 

Grounded-cathode  amplifier  132 

Grounded-grid  amplifier  ....132,  135,  140,  214,  258 

Grounded-grid,  Class  B amplifier  139 

Grounded-grid  r-f  amplifier  626 

Grounded  neutral  wiring  565 

Grounded-screen  r-f  amplifier  629 

Guy  wires,  antenna  449 

H 

Hairpin  coupling  232 

Half-wave  rectifier  689 

Handie-Tolkie,  144  Me 547 

Harmonic  — B.f.o 225 

” Crystal  246 

■■  Def.  of  58 

” Distortion  119 

’’  Music  135 

" Oscillator  247 

” Radiation  262,  373,  725 

Harmoniker  TVI  filter  379 

Hartley  oscillator  240 

Hash  rectifier  679,  694 

Heat  cycle,  resistor  357 

Heot  sink  99,  147 

Heat  sink,  transistor  576,  707 

Heater  cathode  70 

Heising  modulation  294 

Helical  antenna  483 

Henry,  def.  of  37 

Hertz  antenna  426 

Heterodyne  208 

H-F  antennas  459 

High  fidelity  134-138 

” Interference  386 

High-pass  TVI  filter  372 

Holes  in  semi-conductor  91 

Horizontal  directivity  410 

Horn  antenna,  VHF  486 


790 


Hot  cathode  oscillator  580,  676 

Hot  cathode  phase  inverter  116 

Hum,  audio  142 

Hysteresis  loop  — loss  36,  38 


I 

1-f  alignment  ......148,  235 

” Amplifier  210 

” Band-width  graph  534 

" Filter,  lattice  579 

" Noise  limiter  227 

'■  Pass-band  219 

" Rejection  notch  222 

" Shape  factor  219 

" Tuned  circuit  218 

Ignition  noise  528 

Images — Image  intereference — ratio  211,  213,  385 

Impedance  46,  47 

” Antenna  408 

’’  Complex  51 

■■  Coupling  56,  113 

” match,  audio  127 

" reflected  56,  63 

" Resonant  54 

” Screen  circuit  289 

” Surge  424 

" Tank  circuit  262 

" Transformation  63 

” transistor  97 

" Transmission  line  421 

" Triangle  48 

Incident  voltage  732 

Indicator,  Antennascope  738 

Indicator,  selsyn  514 

Indicator  — standing  wave  — twin-lamp  731,  735 

Induced  current  42 

Inductance  37 

" Copacitor  359 

” Cathode  lead  122 

Critical  686 

Lead  — Magnetic  — Mutual  81,  37 

" Parallel  — series  38 

Resistor  357 

" Screen  lead  257 

Induction,  def,  of  42 

Inductive  reactance  46 

Inductive  coupling  271 

Inductive  tuning  617 

Inductor  input  filter  687 

Inductor,  iron  core  38 

Inductor,  time  constant  40 

Infinite  impedance  detector  224 

Initial  condition  voltage  .203 

Injection  voltage  213 

Input  loading  122 

Input  resistance  107,  216,  276 

Insertion  loss,  filter  64 

Instability,  R-F  amplifier  615 

Instability,  static  117 

Insulation,  VHF  479 

Insulator,  def.  of  22,  23 

Integration,  electronic  198 

Integrator  (RC)  59 

Interelectrode  capacitance  76 

Interference  — Broadcast  379 

” Harmonic  373 

’’  Hi-fi  — image  — TV  386,  385,  371 

Interlock,  power  395 

Intermodulation  distortion  136 

Intermodulation  test  145 

Internal  resistance  25 

international  Morse  Code  15 

Interruption  frequency  209 


791 


Interstage  coupling  270,  619,  665 

Intrinsic  semi-conductor 91 

Inverse  current  — Voltage  693 

Inverter,  phase  — voltage  divider  ....1 15,  116,  117 

Ion,  def.  of  — positive  22,  87 

Ionosphere,  absorption  13 

Ionospheric  layers  — propagation  417,  416 

IR  drop  24 

Iron  vane  meter  723 

Isotropic  radiator  410 

J 

Jitter  154,  188 

Johnson  Q-feed  system  446 

“J"  operator  47 

Joule,  def.  of  30,  37 

Junction,  transistor  93 

K 

K,  dielectric  constant  32 

Key  clicks  396 

Keying  — blocked  grid  398 

” Cathode  circuit  397 

” Differential  400 

” Frequency  shift  326 

” screen  grid  399 

Sequence  677 

” Transmitter  395 

Kilocycle  42 

Kinescope  tube  84 

Kirchhoff's  Laws  28 

Klystron  — reflex  81,  82 

L 

Lamb  noise  limiter  227 

Layout,  chassis  726 

Lazy-H  antenna  468 

L/C  ratio  56 

Lead  inductonce  81 

Leakage,  capacitor  34 

Leakage  reactance  63 

Left-hand  rule  35 

Length,  antenna  405 

L-section  filter  64 

Licenses,  amateur  12 

Limiter,  audio  228 

Limiter,  diode  151 

Limiter,  F-M  324 

Limiter,  full-wave  230 

Limiter,  grid  153 

Limiter,  noise  227 

Limiter,  series  diode  185 

Limiter,  TNS  230 

Limiting  circuit 151,  152,  185 

Limiting,  grid  187 

line,  delay  204 

Line  filter  227 

Line  regulation  388 

Line,  Slotted  730 

Linear  amplifier  286 

" 4CX-1000A  641 

■'  Class  B 129 

” SSB 620,  624 

’’  Tuning  287 

Linear  matching  transformer  446 

Linearity  tracer  151 

Link  circuit  611 

Link  coupling  — antenna  272,  452 

Lissajous  figures  144 

litz,  wire  54 

L-network  design  265 

Load  line  74 

Load  line,  transistor  99 


Load  resistance  amplifier  119 

Loaded-Q  57 

Loftin-White  amplifier  118 

Long-wire  antenna  461 

Loop  feedback  158 

Loran,  band  12 

Loudness  control,  audio  140 

Loudspeaker  — def.  of  — response  ..139,  140 

Low-frequency  porosities  366 

Low-pass  TVI  filter  376 

M 

Magic  eye  tube 88,  226 

Magnetic  — Air  gap  38 

” Circuit  — field  — flux  35 

" eddy  current  38 

” hysteresis  loss  38 

“ induction  37 

" left  hand  rule  35 

” Permeability  — reluctance  — saturation  ..36 

Magnetism  — residual  35,  37 

Magnetomotive  force  {M.  M,  F.)  36 

Magnetron  83 

Magnitude,  scalar  48 

Majority  carrier  93 

Marconi  antenna  408,  432,  434,  438 

Mast,  A-frame  448 

Matching  stub,  antenna  445,  503 

Matching  systems  antenna  442 

Matching  transformer  429 

Mathematics,  radio  742 

Maximum  usable  frequency  13,  417 

Maxwell,  def.  of  36 

M-derived  filter  64 

Meosurements  — Antenna  730 

" Bridge  728 

” Circuit  constants  727 

" Current  — voltage  721,  723 

" Frequency  729 

’*  Parallel  wire  line  735 

'*  Power 721,  725 

" Transmission  line  730,  734 

Mechanical  filter  222 

Mechanical  filter  adapter  S.S.B 535,  590 

Megacycle 42 

Megohm  24 

Memory  circuit  204 

Mercury  vapor  rectifier  694 

Meteor  bursts  420 

Meter,  d'Arsonval  - 721 

Meter,  iron  vane  723 

Meter,  multi-range  722 

Meter,  rectifier  724 

Meter,  thermocouple  724,  726 

Mho,  def.  of  74 

Mica  dielectric  32 

Microfarad  30 

Microhenry  37 

Micro-ohm  23 

Middle  C 134 

Miller  effect  107,  1 12,  220 

Miller  feedback  — oscillator  199,  247 

Milliammeter  721 

Millihenry,  def.  of  37 

Mixer  — circuits  212 

’’  Diode  80 

” Noise  212 

" Products,  SSB  — spurious  341,  343 

" SSB  340 

" Stage  211 

” Transistor  100 

" Triode  213 

” Tube  79 


Mobile  — Amplifier,  SSB  620 

Antenna  coupling  520,  524 

Control  circuit  525 

Dynamotor  526 

Equipment  construction  design  ....524,  515 

Filament  regulator  552 

Filter  circuits  518 

Noise  limiter  516 

Noise  sources  529 

Power  supply  521,  697,  519 

Supply,  three-phase  620 

Mode  of  resonance  232 

Modulation  239 

Amplitude  282,  647 

Automatic  control  648 

Bandwidth  283 

Cathode  297 

Class  B 294 

Constant  efficiency  286 

Distortion  307 

Frequency  312 

Grid  252,  288 

Heising  294 

Index,  F-M  314 

Pattern,  oscilloscope  146 

Percentage  284 

Phase  315,  319 

Plate  251,  293,  295 

Screen  288 

Suppressor  292 

Transformer  295 

Variable  efficiency  285 

Velocity  81 

Modulator  — Adjustment  654 

Balanced  — S.S.B 331,  338,  343 

Bias-shift 307 

Cathode  follower  290 

Class  B 125 

Construction  650 

Crosby  590 

Diode  332 

Matching  - — impedance  match  125,  127 

Reactance  tube  315 

SSB  340 

Tetrode  647 

Zero  bias  662 

Molecule  21 

Monitor  oscilloscope  147 

Morse  Code  15 

Mu  factor  (^t)  78 

Multee  antenna  440 

Multi-band  antenna  436,  514 

Multiplication,  electronic  198 

Multiplier,  frequency  258 

Multiplier  resistor  722 

Multi-range  meters  722 

Multivibrator  — circuits  102,  154,  188 

Multivibrator,  free  running  155 

Multi-wire  doublet  429,  442 

Music,  def.  of  134 

Music  systems  — scale  138,  1 35 

Mutual  conductance  73 

Mutual  coupling  218 

Mutual  coupling,  antennas  475 

Mutual  inductance  37 

Mycolex,  dielectric  32 


N 


Narrow-band  F-M  315 

NBS  bridge-T  oscillator  192 

Negative  feedback  loop  140 

Negative  peak  clipping  648 

Negative  screen  current  641 


792 


Networks,  L and  Pi  265 

Network,  phasing  570 

Neutralization  — Amplifier  252 

" Capacitance  107 

" Capacitor,  VHP  672 

' Grid  -253 

” Procedure  279 

” R-F  276 

" R-F  amplifier  616,  639 

” Shunt  254 

’■  Test  278 

” Transistor  100 

Neutralizing  procedure  255 

Nl  (ampere  turns)  62 

Noise  — Factor  122 

" Check,  mobile  528 

" Generator,  silicon  crystal 740 

” Limiters  — mobile  227,  516 

” Mixer  212 

” Regulator  — sources  528 

" Suppression  227,  527 

” Thermal  agitation  188 

” Voltage  121 

" Wow  and  flutter  135 

Nomenclature,  component  530 

Nonconductor  22 

Nondirectional  VHF  antenna  482 

Non-linear  distortion  135 

Non-linear  function  202 

Non-resonant  transmission  line  421 

Nonsinusoldal  wave  58 

N-P-N  transistor  92 

O 

Octave,  def.  of  135 

Oersted,  def,  of  36 

Ohm  — Ohm's  Law  23,  24 

Ohmmeter,  low  range  723 

Ohm's  Law  (a.c.)  45 

Ohm's  Law  (complex  quantities)  49 

Ohm's  Law  for  magnetic  circuit  36 

Ohm’s  Law  (resonant  circuit)  54 

Ohms-per-volt  722 

Omega,  def.  of  44 

One-shot  multivibrator  189 

On-off  circuits  195 

Open  wire  line  421 

Operating  desk,  construction  of  724 

Operational  amplifier  199 

Orbital  electron  21 

Oscillation,  parasitic  279,  365,  616 

Oscillator  — Beat  224 

" Blocking  102,  155,  156,  189 

" Bridge-T  192 

" Circuits  250 

" Clapp  — Colpitts  241 

" Code  practice  19 

" Crystal  244 

" Dynatron  242 

’’  e.  c,  o 241 

'■  Franklin  243 

” Free  running  141 

" Harmonic  247 

” Hartley  240 

" High  stability  604 

" Hot  cathode  580,  676 

" Keying  249 

" Miller  247 

" NBS  bridge-T  192 

” Overtone  249 

’’  Phase  shift  157,  191 

’*  Pierce  247 

" RC 157 


Oscillator  — Relaxation  102 

” t.p.t.g 241 

" Transistor  101 

" Transistron  243 

” V.F.O 244 

” Wien-bridge  157,  191 

Oscilloscope  138 

'•  Circuit  142 

” Linearity  tracer  151 

" Lissajous  figures  144 

" Modulation  pattern — phase  patterns  146,  145 

" Monitor  147,  149 

" Sideband  measurements  150 

" Trapezoidal  pattern — wave  pattern  146,  148 

Output,  peak  power  124 

Overloading,  TV  receiver  371 

Overtone  crystals  249 

Overtone,  music  135 

Oxide  filament  67,  69 

P 

Paper  dielectric  32 

Parallel  circuit  — resistance  25 

" Diode  limiter  186 

” Feed  273 

" Resonance  54 

" Wire  line  measurements  735 

Parasitic  antenna,  design  chart  498 

" Antenna,  VHF  488 

" Beam  design  494 

" Check  for  368 

" Coupling  364 

” Element,  antenna  497 

” Oscillation  279,  365,  616 

” Resonance  364 

Pass  band,  1-F  219 

Pass  band,  mechanical  filter  223 

Pass  band,  SSB  filters  336 

Potterns,  antenna  404,  460 

PE-103  dynamotor  526 

Peak  amplifier  current  125 

Peak  current  (a.c.)  45 

Peak  envelope  power,  SSB  327 

Peak  limiter  185 

Peak  noise  limiter  227 

Peak  power  output  124 

Peoked  wave  59 

Pentode  amplifier  120 

Pentode  tube  77 

Period,  sound  1 34 

Permeobility  36 

Phantom  signal  382 

Phase  — angle  (a.c.)  46 

” Angle  — difference  146,  145 

" Distortion  109,  135 

" Inverter  1 1 5 

" Modulation  315,  319 

" Shift,  clipping  301 

" Shift,  feedback  193 

" Shift,  oscillator  157,  191 

Phasing,  antenna  466 

Phasing  generator,  SSB  337 

Phasing  networks,  audio  338 

Phasing  network,  SSB  570 

Phonograph  reproduction  136 

Photoelectric  emission  67 

Pickups  137 

Pickup,  spurious  384 

Pierce  oscillator  247 

Pi-network  amplifier  617 

Pi-network  chart  — design  267,  265 

Pi-network  coupling  453 

Pilot  carrier,  SSB  327 


793 


Pitch,  sound  134 

Plate  current  flow  259 

Plate  detector  224 

Plate  efficiency  amplifier  129 

Plate  modulation  251,  293 

Plate  resistance  73,  74 

P-N-P  transistor  91 

Point  contact  transistor  92 

Polar  notation  48 

Polarization,  antenna  — VHP  404,  479 

Polyphase  rectifier  693 

Potential  difference  22 

Potential,  electrode  106 

Potential  energy  30 

Power  amplifier  design  610 

Power  amplifier  triode  118 

Power,  gain  94 

Power  gain  antenna  405 

Power  gain  SSB  328 

Power  measurement  721,  725 

Power  interlock  395 

Power-line  filter  227 

Power  loss,  VHP  663 

Power,  resistive  29 

Power  supplies  684 

” A.C.-D.C 694 

” Mobile  697 

” Oscilloscope  141 

” Regulated  711 

" Screen  269 

" Selenium  561 

Three-phase  521,  620 

" Transistorized  703 

’’  V.P.O 604 

” Voltage  multiplier  695 

Power  system,  primary  387 

Power  transformer  709 

Power  transistor  99 

Powerstat  auto-transformer  390 

Preamplifier,  hi-fi  138 

Pre-emphasis,  recording  137 

Preselector  214 

Primary  transformer  61 

Printed  circuit  amplifier  575 

Probe,  R.F.  (v.t.v.m.)  133 

Product  detector  352,  237 

Propagation  413-420 

Pulsating  alternating  current  45 

Pulse-repetition  frequency  103,  190 

Push-pull  amplifier  612,  121 

Push-pull  tetrode  amplifier  644 

Push-pull  transformer  113 

Push-pull  tripler  — doubler  260 

Push-to-talk  circuit  526 

Q 

Q amplifier  tank  128 

Q antenna  409 

Q coils  360 

Q def.  of  54 

Q loaded  — unloaded  57 

Q multiplier  236 

Q tank  circuit  261 

Q transformer  429 

Q tuned  circuit  21  6 

Quad  antenna  471 

Quadrant,  sine  43 

Quantity,  complex  49 

Quench  oscillator  209 

R 

Radian,  def.  of  43 

Radiation,  angle  of  411,  460,  478 


794 


Radiotion,  def.  of  403 

Radiotion,  harmonic  373,  725 

Radiation  pattern,  distortion  478 

Rodiation  resistance  404,  495 

Radiator,  isotropic 410 

Radiator  cross-section,  VHF  479 

Radio  frequency,  def.  of  42 

Radio  mathematics  742 

Radio  propagation  413 

Radio  teletype  326 

Romp  function  203 

Ratio  detector  323 

Ratio,  image  2 1 3 

RC  amplifier  109 

RC  audio  network  139 

RC  differentiator  — integrator  59 

RC  oscillator  — circuits  157,  191 

RC  time  constant  60 

RC  transient  38 

Reactance,  antenno  408 

Reactance,  capacitive  — inductive  — net  46 

Reactance,  leakage  63 

Reactance,  resonance  47 

Reactance  tube  modulator  315 

Read-out  201 

Receiver,  olignment  148,  235 

Receiver,  amateur  band  540 

Receiver,  superheterodyne  210 

Receiver,  transistor  103,  533 

Receivers,  UHP  231 

Reception,  frequency  modulation  321 

Reception,  mobile  515 

Reception,  SSB  351 

Recording,  constant  amplitude,  velocity  ....136,  137 

Recording,  crossover  point  136 

Recording,  high  fidelity  136 

Recording,  pre-emphasis  — RIAA  curve  137 

Rectification,  stray 383 

Rectified  A.C 45 

Rectifier  - — bridge  — full  wave  — half  wave  ....689 

” Hash  679,  694 

” Mercury  vapor  polyphase  693,  694 

” Selenium  695 

” Silicon  592,  696 

” Type  meter  724 

” Vacuum  693 

” v.t.v.m 133 

Reflected  impedance  57,  63 

Reflected  voltage  732 

Reflectometer,  coaxial  732 

Regeneration  208 

Regulated  power  supply  711 

Regulation,  power  line  388 

Regulation,  voltage  686 

Regulator,  filament  552 

Regulator  noise  528 

Regulator  tube  (VR)  709 

Regulator,  voltage 687 

Rejection  notch,  l-P  222 

Rel,  def.  of  36 

Relaxation  oscillator  102,  188 

Reluctance,  def.  of  36 

Remote  cut-off  tube  78 

Residual  magnetism  37 

Resistance:  23 

” capacitance  filter 687 

*'  Dynamic  95 

" gain  94 

” Ground  410 

” Input  107,  216 

” Internal  — parallel  — series  25 

" load  119 

” plote  73,  74 


Resistance  — Radiation  404 

Resistive  power  29 

Resistivity,  table  of  - - 23 

Resistor,  bleeder  27,  709 

Resistor,  characteristics  of  356 

Resistor  color  code  531 

Resistor,  equalizing  34 

Resistor,  inductance  of 357 

Resisto**  multiplier  722 

Resistor,  typical  24 

Resonance:  47 

" Antenna  404 

" Current  56 

" Curve  — parallel  — Series  53,  54 

” Filter  688 

" Impedance  — Q 54 

” Mode  232 

" Oscilloscope  pattern  149 

" Parasitic  364 

" Speaker  140 

Resonant  cavity 232 

Resonant  circuit  53,  54 

Resonant  transmission  line  424 

Response,  audio  112,  136 

Return  trace  140 

R-F  alignment  236 

R-F  amplifier:  213,  251 

" 4CX-1000A 640 

" Cathode  driven  626 

" Construction  616 

" General  purpose  635 

" Grounded  grid  — grounded  screen  626,  629 

" Inductive  tuning  617 

Instability  — neutralization  ....615,  616,  639 

Oscillation  616 

Pi-network  617 

" Push-pull  612 

" Self-neutralization  616 

Tetrode  614 

R-F  chokes  272,  361 

R-F  feedback  274 

R-F  ground  362 

R-F  interstage  coupling  619 

R-F  shielding  362 

Rhombic  antenna  464,  486 

Rhumbatron  cavity  232 

RIAA  equalizer  curve  137 

Ribbon  TV  line  478 

Ring  diode  modulator  333 

Ripple  factor,  filter  688 

Ripple  voltage  686 

RL  circuit  40 

RL  transient  38 

RLC  circuits  48 

Root  mean  square  (a.c.)  45 

Rotary  beam  antenna  494 

Rotator,  antenna  512 

Ruggedized  tube  88 

S 

Safety  bleeder 395 

Safety  precautions  393 

Saturation  current  72 

Saturation,  magnetic  36 

Sawtooth  wave  59,  140 

Scalar  notation  48 

Scale,  musical  135 

Scatter  signals  419 

Screen,  CRT  87 

Screen  current,  negative  641 

Screen  grid  keying  399 

Screen  grid  tube  77 

Screen  lead  inductance  257 


Screen  modulation  288 

Screen  supply  269 

Secondary  emission  71,  77 

Secondary  transformer  61 

Selective  fading,  SSB  331 

Selectivity,  arithmetical  211 

Selectivity  chart 342 

Selectivity  control,  l-F  221 

Selectivity,  mobile  reception  517 

Selectivity,  resonant  circuit  54 

Selectivity,  tuned  circuits  341 

Selenium  power  supply  561,  669 

Selenium  rectifier  695 

Self-inductance  37 

Self-neutralization  amplifier  616 

Self-neutralization,  VHF  664 

Selsyn  indicator 514 

Semi-conductor  90,  21 

Semi-conductor  rectifier  695 

Sequence  keying  677 

Series-cathode  modulator  297 

Series  circuit  24 

Series-derived  filter  64 

Series-diode  limiter  151,  185 

Series  feed  273 

Series-feed  amplifier  612 

Series-parallel  circuit  25 

Series  resistance  — resonance  25,  53 

Shape  factor,  l-F  219 

Shielding,  R-F  362 

Shot  effect  123 

Shunt  derived  filter  64 

Shunt  loading,  a.v.c 226 

Shunt  neutralization  254 

Sidebands,  def.  of  282 

Signal,  error  159,  193 

Signal,  phantom  382 

Signal-to-distortion  ratio  — SSB  151,  344 

Signol-to-noise  rotlo  122 

Silicon  crystal  noise  generator  740 

Silicon  rectifier  592,  696 

Sine  wave  42,  43 

Single-ended  amplifier  611 

Single  sideband  (SSB)  : 285,  329,  327 

" Amplifier,  4CX-1000A  640 

Envelopes  330 

Envelope  detector  151 

” Equipment  567 

” Filter  327 

" Filter  system  590 

Grounded  grid  amplifier  626 

Grounded  screen  amplifier  629 

Jr.  exciter  346 

" Linear  amplifier  620 

Linear  amplifier,  kilowatt  635 

" Linear  operation  138 

Measurements  150 

” Reception  351 

” Transmission  327 

” Transmitter  adjustment  573 

Single  signal  reception  222 

Single  swing  oscillator  189 

Single-wire  antenna  tuner  456 

Single-wire  feed,  antenna  441 

Single-wire  feeder  430 

Six-shooter  antenna  472 

Skeleton  VHF  antenna  481 

Skin  effect  54 

Skip  distance  418 

Sky  wave  414 

Sleeve  antenna  480 

Slide-wire  bridge  728 

Slope  detection  210 


795 


Slotted  line  730 

S-meter  circuits  226 

Soldering  techniques  727 

Sound  in  air  1 34 

Space  charge  71,  78 

Space  wave 414 

Speaker  response  140 

Speaker  tweeter  139 

Specific  resistance  23 

Speech  amplifier  construction  655 

Speech  clipper  filter  304 

Speech  clipping  124,  291,  300,  648 

Speech  filter,  high  level  305 

Speech  waveforms  285,  294,  301,  648 

Splatter  suppressor  302,  647 

Sporadic-E  propagation  418 

Spotter,  frequency 730 

Spurious  emission  — spurious  pickup  383,  384 

Spurious  products,  mixer  343 

Square  wave  — square  wave  test  58,  61,  62 

Stacked  antenna  — stacked  dipole  498,  468 

Standard  frequency  729 

Standard  pitch  135 

Standing  wave  403,  423 

Standing-wave  indicator  731 

Static,  corona  529 

Static,  wheel  528 

Steering  diode 1 03 

Step-by-step  counter  156,  190 

Sterba  antenna  469 

Storage  time,  diode  103 

Stray  capacitance  218 

Stub  match,  antenna  503 

Summation  voltage  197 

Summing  amplifier  201 

Sunspot  cycle  419 

Superheterodyne  receiver  210 

Super-imposition  tuning  553 

Super-regenerative  detector  209 

Supply,  a.c.-d.c 588 

Suppression,  audio  306 

Suppression  circuits,  porosities  366 

Suppressor  77 

Suppressor  modulation  292 

Suppressor,  splatter  302 

Surface  wave  414 

Surge  Impedance 424 

Susceptance,  def.  of  52 

Sweep  frequency 140 

Switching,  transistor  action  704 

SWR  bridge  458,  736 

Symmetry,  amplifier  612 

Synchronization,  generator  173 

Synchronizing  voltage  172 

Synchronous  detection,  DSB  353 

T 

Tank  capacitance  261 

Tank  circuit  — efficiency  55,  57,  58 

Tank  circuit  impedance  — loading  262,  264 

Tank  circuit  Q 1 28 

Technician  class  amateur  license  12 

Teletype,  radio  326 

Television  interference  371 

Temperature  coefficient  32 

Ten-A,  SSB  exciter  347 

Ten-meter  transceiver  559 

Terman-Woodyard  amplifier  298 

Termination,  transmission  line  425 

Termination,  VHP  rhombic  487 

Test  equipment  721 

Tetrode  amplifier,  push-pull  644 

Tetrode  modulator  647 


Tetrode  neutralization  256 

Tetrode  r-f  amplifier  614 

Tetrode,  screen  safety  circuit  679 

Tetrode,  self  neutralization  664 

Tetrode  tube  77 

Thermal  agitation  noise  188 

Thermionic  emission  67 

Thermocouple  meter  724,  726 

Three-phase  power  supply  521,  620 

Threshold  voltage  696 

Thyratron  tube  87 

Time-base  generator  139 

Time  constant  38,  60 

Time  sequence  keying  400 

T-match  antenna  501 

T-match  system  444 

TNS  limiter  230 

Tools,  radio  721 

T.P.T.G.  oscillator  241 

Trace  — cathode  ray  86,  140 

Tracking  capacitor  216 

Transceiver,  28  me 559 

Transceiver,  50  me 552 

Transceiver,  VHP  578 

Transconductance  74,  79 

Transducer  — pickup  138,  137 

Transformation,  impedance  63 

Transformation  rotio  62 

Transformer  — ampere-turns  61,  62 

” Antenna  match  502 

’*  Auto  63 

” Coupling  1 1 3 

" l-F  211 

” Linear  matching  446 

■’  Matching  — impedance  match  429,  63 

" Modulator  295 

**  Power  709 

" Vibrator  588 

Tronsformerless  power  supply  694 

Transient  circuit  (a.c.)  59 

Transient  distortion  135 

Transient,  RC,  RL  — transient  wave  38,  58 

Transistor  90,  92 

" Bias  96 

" Code  oscillator  19 

” Complementary  circuit  97 

” Drift  93 

” Equivalent  circuit  95 

” Heat  sink  576,  707 

” Impedance  97 

” Junction  91 

” Mixer  100 

■'  Multivibrator  102 

" Oscillator  101 

” Point-contact  95 

” Power  99 

” Receiver  533 

" SSB  equipment  574 

" Switching  action  704 

Transistorized  mobile  supply  703 

Transit  time  81 

Transitron  oscillator 243 

Transmission,  duplex  597 

Transmission  line  chart  422 

” Circuits  231 

” Coaxial  423 

” Discontinuities  425 

” Impedance  421 

” Measurements  730,  734 

’*  Resonant  — non-resonant  424,  421 

" Termination  425 

" VHP 478 

Transmitter  — Construction  663 


796 


Transmitter  — Control  circuits  391 

" Design  356 

Ground  394 

" Keying  395 

" Oscilloscope,  monitor  of  147 

Receiver,  VHF  597 

Trapezoidal  pattern,  oscilloscope  146 

Trap-type  antenna  514 

Travelling  wave  tube  84 

Travis  discriminator  322 

Triode  amplifier  110 

Triode  mixer  213 

Triode  power  amplifier  118 

Triode  tube  72 

Tripler,  push-pull  260 

Triplex  antenna  475 

Tropospheric  propagation  4T5 

T-section  filter  64 

Tube,  vacuum  67 

Tube,  VHF  design  232 

Tube,  water  cooled  (4W-300B)  620 

Tubular  transmission  line  422 

Tuned  circuit,  coaxial  232 

Tuned  circuit,  l-F  21  8 

Tuned  circuit,  r.f  amplifier  121 

Tuned  circuit  Q 216 

Tuned  circuits,  selectivity  341 

Tuner,  antenna  456 

Tuning,  bandspead  217 

Tuning  Indicators  226 

Tuning,  inductive  617 

Tuning,  super-imposition  553 

Turnstile  antenna  480 

7VI  filters  373,  378 

TVI-proof  enclosures  — openings  724,  725 

TVI  suppression  375 

TVI-type  filter  64 

Tweeter  speaker  139 

Twin  lamp 735 

Two-band  Marconi  antenna  438 

U 

Ultra-linear  amplifier  142,  146 

Unidirectional  antenna  504 

Unl-potentlal  cathode  70 

Unity  coupling  271 

Unloaded  Q 57 

V 

Vacuum  capacitor  360 

Vacuum  tube:  67 

” Amplification  73,  74 

” Beam  power  78 

Cathode  ray  84 

Classes  — classification  107,  87 

Conductance  73 

■'  Constant-current  curve  127 

' Diode  71 

Equivalent  noise  resistance  123 

" Foreign  89 

' Gas  87 

” Input  loading  122 

” Klystron  81 

Load  line  74 

" Magic  eye  88 

Magnetron  83 

’ Mixer  79 

Operation  75 

" Parameters  106 

" Pentode  77 

Plafe  resisfance  73,  74 

” Polarity  reversal  76 

” Remote  cutoff  78 


Vacuum  — Shot  effect  123 

" Space  charge  89 

" Tetrode  77 

” Thyratron  87 

” Travelling  wave  84 

" Triode  72 

*'  Upper  frequency  232 

" Variable  mu  (^t)  124 

*'  VHF  80 

" Voltage  regulator  87 

" Voltmeter  130,  724 

V-Antenna  462 

Variable-mu  (fi)  tube  124 

Variable  reluctance  pickup  137 

Variac  autotransformer  390 

Vector,  sine-wave  43 

Vehicular  noise  suppression  527 

Velocity  modulation  81 

Vertical  antenno  430 

Vertical  directivity  411 

VFO,  construction  676 

VFO,  high-stability  604 

VHF  Amplifier  214 

" Antennas  477 

Antenna  polarization  479 

” Antenna  relay  478 

” Bands,  characteristics  14 

” Corner  reflector  antenna  485 

” Coupling,  interstage  665 

” Definition  of  477 

” Discone  antenna — ground-plane  antenna  481 

" Handie-talkie  547 

” Helical  antenna  — Horn  antenna  ....483,  486 

" Multi-element  antenna  488 

” Neutralization  capacitor  672 

'*  Nondirectional  antenna  482 

” Parasitics  366 

” Power  loss 663 

” Radiotlon  angle  — radiation  pattern  ....478 

“ Rhombic  antenna 486 

’’  Screen  antenna  491 

” Self-neutralization  664 

” Sleeve  antenna  480 

’’  Transceiver  552,  578 

" Transmission  line  478 

'*  Transmitter-receiver  597 

" Turnstile  antenna  480 

" Wavelength  table  479 

Vibrotor  power  supply  588 

Vibrator,  split-reed  698 

Video  amplifier  112,  128 

Volt,  def.  of  22 

Voltage  — Amplifier  110 

” Avalanche  697 

’’  Break  203 

” Breakdown  32 

” Decay  40 

” Divider  26,  52 

” Divider,  phase  inverter  117 

” Drop,  summation  28 

" Gradient  39 

” Incident  732 

” Injection  213 

” Initial  condition  203 

” Instontaneous  44 

" Inverse  693 

" Measurement  721,  723 

" Multiplier  power  supply 695 

" Noise  121 

’*  Output  685 

” Reflected  732 

” Regulation  686,  687 

” Regulator  tube  87,  709 


797 


Voltage  — Resonant  54 

" Ripple  686 

" Summation  197 

” Synchronizing  141 

" Threshold  696 

Voltmeter  722 

Voltmeter^  A-C  V-T  724 

Voltmeter,  diode  724 

Voltmeter,  vacuum  tube 130,  724 

Volt-ohmmeters  722 

Volume  compression  648 

W 

Wagner  ground  729 

Water  cooled  tube  {4W-300B)  620 

Watt,  def.  of  29 

Wave  — Carrier 207 

" Ground  414 

” Harmonic  — nonsinusoldal  58 

” Pattern,  oscilloscope  148 

” Peaked  — sawtooth  59 

" Sky  — space  414 

''  Square  — transient  58 

" Surface  414 

Waveform  59,  143 

Waveform,  speech  285,  301,  648 

Wavelength,  def.  of  405 

Wavelength  table,  VHP  479 


Wave-shaping  circuits  

Weston  cell  

Wheatstone  bridge  

Wheel  static  

Wien-bridge  oscillator  

Williamson  amplifier  

Wire,  litz  

Wiring  hints  

Wiring  technique,  audio  

Workshop  practice  — layout 

W8JK  antenna  

WWV  transmissions 


X 

X-array  antenna  

Y 

Yogi,  adjustment  

Y,  (admittance,  def.  of)  

Yogi  antenna  — feed  systems 

Yagi  antenna,  VHP  

Yogi,  construction  

Yoke  match  


z 

Zeppelin  antenna  

Zero-bias  modulator  


PREPARES  YOU  FOR  THE  F.C.C. 
RADIOTELEPHONE  LICENSE 
EXAMINATION 


Now,  in  one  convenient  volume,  complete  study-guide  ques- 
tions with  clear,  concise  answers  for  preparation  for  oil  U.S.A. 
commercial  radiotelephone  operator's  license  examinations. 

CONTAINS  FOUR  ELEMENTS: 

I.  QUESTIONS  ON  BASIC  LAW 

II.  BASIC  OPERATING  PRACTICE 

III.  BASIC  RADIOTELEPHONE 


1 85 

24 

728 

528 

157,  191 

141 

54 

531 

142 

.720,  729 

473 

729 


469 


509 

5 2 

494,  498 

488 

506 

498 


.427 

662 


Bookstores:  Order  from  Baker  & Taylor  Co.,  Hillside,  N.J. 


IN  TWO  VOLUMES 

This  set  of  conversion  data  hos  become  standard  for  most 
commonly  used  items  of  surplus  electronic  equipment.  All 
conversions  shown  are  practical  and  yield  o useful  item  of 
equipment;  oil  have  been  proven  by  testing  on  severol  units. 

VOLUME  I 


6C-221  Frequency  Meter 
BC-342  Receiver 
BC-348  Receiver 
BC-312  Receiver 
BC-412  Oscilloscope  os  o 
test  scope  or  as  o tele* 
vision  receiver. 

BC-645  420  - Me.  Tronsmit* 
ter/ Receiver 

BC-453A  Series  Receivers 
BC-457A  Series  Tron'.mifters 
SCR-522  144-Mc.  Transmit- 
ter / Receiver 

TBY  Transceiver  with  Xtal 
Control 

PE-103A  Dynamotor 
BC-1068A  V-h-f  Receiver 
Electronics  Surplus  Index 
Cross  Index  of  VT-Number 
tubes 

Order  Book  No.  EAE-SM1 


VOLUME  tl 

ARC-5  and  BC-454  Receivers 
for  28  Me. 

ARC-5  ond  BC-457  Tx  for 
28-Mc.  Mobile 
ART-13  and  ATC  Xmitter 
Surplus  Beam  Rototing 
Mechanisms 

Selenium-Rect.  Power  Units 
Hi-Fi  Tuner  from  BC-946B 
Receiver 

ARC-5  V-h-f  Transmitters 
GO-9  and  TBW  Xmitters 
9-W  Amplifier  from  AM-26 
TA-12B  & TA-12C  Xmitters 
AVT-1  12A  Aircroft  Xmitter 
BC-375  & BC-191  Xmitters 
Model  LM  Freq.  Meter 
Primary  Power  Requirements 
Chort 

ARB  Recr.  Diogrom  Only 
Order  Book  No.  EAE-SM2 


$2.50  for  either  volume 


BUY  FROM  YOUR  FAVORITE  DISTRIBUTOR 

Of  obovc  price  or  odd  1 0 on  direct  motl  orders  to: 
EDITORS  and  ENGINEERS,  Ltd. 
Summerlond,  Coiifornio 


Bookstores:  Orcier  from  Baker  & Taylor  Co.,  Hillside,  N.J. 


WORLD'S  RADIO  TUBES 

(RADIO  TUBES  VADE  MECUM) 


Characteristics  of  all  existing  radio  tubes 
made  in  all  countries.  The  worldV  most 
authoritative  tube  book.  All  types  classified 
numerically  and  alphabetically.  To  find  a 
given  tube  is  only  a matter  of  seconds!  Also 
complete  section  on 

base  connections.  $5.00 


WORLD'S  EQUIVALENT  TUBES 

1 (EQUIVALENT  TUBES  VADE  MECUM) 


All  replacement  tubes  for  a given  type,  both 
exact  and  near-equivalents  (with  points  of 
difference  detailed).  From  normal  tubes  to 
the  most  advanced  comparisons.  Also  military 
tubes  of  7 nations,  with  commercial 
equivalents.  Over  43,900  comparisons!  $5.00 


WORLD'S  TELEVISION  TUBES 

i (TELEVISION  TUBES  VADE  MECUM) 


Characteristics  of  all  TV  picture  and  cathode 
ray  tubes.  Also  special  purpose  electronic 
tubes.  Involuable  to  technicians  and 
all  specialists  In  the  electronic  field.  $5.00 


ETTL 

BUY  FROM  YOUR  FAVORITE  DISTRIBUTOR 


Qt  obovc  price  or  odd  1 0 on  direct 


]j|  orders  to; 


EDITORS  and  ENGINEERS,  Ltd. 
Summerland,  California 


Bookstores:  Order  from  Baker  & Taylor  Co,,  Hillside,  N.J.